Recycling of Nickel Metal Hydride (Nimh) Batteries: Characterization and Recovery of Nickel, Ab Alloy and Cobalt

THESIS FOR THE DEGREE OF LICENTIATE OF ENGINEERING
Recycling of nickel metal hydride (NiMH) batteries

Characterization and recovery of nickel, AB5 alloy and cobalt
Filip Holmberg
Industrial Materials Recycling

Department of Chemistry and Chemical Engineering
CHALMERS UNIVERSITY OF TECHNOLOGY
Gothenburg, Sweden 2017
Characterization and recovery of nickel, AB5 alloy and cobalt
FILIP HOLMBERG
© FILIP HOLMBERG, 2017
Licentiatuppsatser vid Institutionen för kemi och kemiteknik

Chalmers tekniska högskola.
Nr 2017:09

Chalmers University of Technology
SE-412 96, Gothenburg
Sweden
Telephone: +46 (0) 31-772 1000
filip.holmberg@chalmers.se
Cover:
The cover image shows SEM micrographs of three components used in nickel metal hydride batteries,
in the following order from back to front: nickel hydroxide particle (cathode), conductive metallic nickel
powder (cathode and anode) and spent hydrogen storage alloy (anode).
Chalmers Reproservice
Gothenburg, Sweden, 2017
ii
FILIP HOLMBERG

Chalmers University of Technology
ABSTRACT
Nickel metal hydride (NiMH) batteries are used today for applications that can assist in the adaptation
toward carbon-neutral energy sources (i.e. hybrid vehicles and smart grids). Recovery of metals such as
nickel, cobalt and rare earth elements (REEs) from discarded NiMH batteries is important for economic
and/or technological reasons. Recirculation of valuable materials back into society can be achieved by
the production of pure concentrates of elements from waste, or by regenerating and reusing materials.
Recycling of NiMH batteries can be done through pyro- and hydrometallurgical methods, which
produces pure and valuable metal fractions. The current work focuses on the separation of three
different types of material from NiMH waste: (1) metallic nickel powder, (2) reusable hydrogen storage
alloy (i.e. AB5) and (3) cobalt-enriched surface compounds.
Bipolar NiMH batteries were investigated in this work, which are different in design compared to
previous investigations on recycling of batteries. In these batteries the conductive nickel network is
made up of nickel powder, instead of conventional nickel foam grids.
Previous investigations suggest that spent electrode materials are not completely degraded. Reusing
parts of the discarded NiMH battery has however not yet been commercially realized. For the current
work spent NiMH batteries were characterized separately. Corrosion products form on the anode
particles during battery cycling, and in this work rare earth hydroxides (i.e. REE(OH) 3) were identified.
The hydrogen storage alloy can potentially be reused if corrosion products are removed.
In the current work it was re-confirmed that nickel, even in powder form, can be recovered by
dissolution of the waste matrix using hydrochloric acid. Cyanex 923 was briefly considered for treating
the leachate solution. The extraction of REEs and aluminum from the leachate was affected by the
amount of acid present, which was attributed to preferential extraction of acid. Controlling the acid in
the dissolution step is therefore important in order to further treat leachate solutions.
The relatively novel recycling approach of recovering AB5 alloy using carboxylic (malonic, maleic, acetic
and citric) acid was considered. An interesting result was that REE(OH) 3 could be removed using acetic
and citric acid, without dissolving much of the remaining spent AB5 alloy. The practical uses were
otherwise limited due to similar dissolution rates of the AB5 alloy and other waste matrix.
A novel recycling approach of using ascorbic acid was suggested and tested. Separation of the cobalt
layers from the cathode hydroxide particles is possible because these cobalt layers have different
oxidation states (higher than 2) compared to the waste matrix. The ascorbic acid treatment does not
dissolve the anode material. Ascorbic acid can therefore potentially be used as a pre-step to chemically
separate the anode material from the cathode material in a mixed electrode waste flow.
iii
LIST OF PUBLICATIONS
This thesis is based on the work contained in the following paper:
Paper I:
Holmberg F., Gutknecht T., Cao, Y., Ekberg C., Steenari B.M. Selective dissolution of cobalt from spent
nickel metal hydride (NiMH) batteries using ascorbic acid, Manuscript.
Contribution Report:
Major experimental work, data treatment, evaluation and major part of writing the manuscript
iv
Table of contents
ABBREVIATIONS ...................................................................................................................................... vii
1. INTRODUCTION .................................................................................................................................... 1
2. BACKGROUND ...................................................................................................................................... 1
2.1 Economic and environmental values of metals in nickel metal hydride (NiMH) batteries ............ 1
2.2 Fundamentals of the NiMH battery ............................................................................................... 2
2.2.1 Hydrogen storage AB5 alloy used in the anode ....................................................................... 3
2.2.2 Nickel hydroxide Ni(OH)2 used in the cathode ........................................................................ 4
2.2.3 Metallic powders nickel and cobalt ........................................................................................ 6
2.3 Degradation mechanisms of battery cells and materials ............................................................... 6
2.3.1 Loss of electrolyte ................................................................................................................... 7
2.3.2 Pulverization ........................................................................................................................... 7
2.3.3 Corrosion ................................................................................................................................ 7
2.3.4 Cross contamination between cathode and anode ................................................................ 8
2.4 Hydrometallurgical recycling methods to treat NiMH waste ......................................................... 9
2.4.1 Dissolution .............................................................................................................................. 9
2.4.2 NiMH metal separation by solvent extraction ...................................................................... 10
2.4.3 Recovery of metals ............................................................................................................... 11
3. THEORY .............................................................................................................................................. 11
3.1 Oxidation of metals ...................................................................................................................... 11
3.2 Recovery of nickel ........................................................................................................................ 12
3.2.1 Recovery of metallic nickel using hydrochloric solutions ...................................................... 12
3.2.2 Distribution into the organic phase in solvent extraction ..................................................... 13
3.2.3 Extraction of chloride complexes with Cyanex 923 .............................................................. 14
3.3 Recovery of reusable anode using carboxylic acids ..................................................................... 14
3.4 Recovery of cobalt using ascorbic acid ........................................................................................ 16
4. EXPERIMENTAL ................................................................................................................................... 18
4.1 Experimental outline.................................................................................................................... 18
4.2 Battery disassembly ..................................................................................................................... 19
4.3 Characterization........................................................................................................................... 20
4.4.1 Dissolution rates of metals at pH 1 ....................................................................................... 21
4.4.2 Nickel dissolution rates at 1-8 M HCl .................................................................................... 22
4.4.3 Leachate treatment with solvent extraction ......................................................................... 22
4.5 Recovery of anode material (including AB5 alloy) ........................................................................ 23
4.5.1 Dissolution experiments using carboxylic acids .................................................................... 24
4.6.1 Dissolution experiments using ascorbic acid......................................................................... 25
v
5. RESULTS AND DISCUSSION ................................................................................................................. 26
5.1 Characterization........................................................................................................................... 26
5.2.1 Effect of acidity on nickel recovery ....................................................................................... 31
5.2.2 Extraction of metals using Cyanex 923 and the effect of acidity .......................................... 34
5.3 Recovery of reusable anode using carboxylic acids ..................................................................... 36
5.3.1 Chemical separation of anode and cathode ......................................................................... 36
5.3.2 Separation of corrosion products from AB5 hydrogen storage alloy..................................... 39
5.4.1 Cathode material .................................................................................................................. 42
5.4.1.1 Leaching of cathode ....................................................................................................... 42
5.4.1.2 Characterization of undissolved cathode materials ....................................................... 45
5.4.2 Anode material ..................................................................................................................... 48
5.4.2.1 Leaching of anode .......................................................................................................... 48
5.4.2.2 Characterization of undissolved anode material ........................................................... 49
6. CONCLUSIONS .................................................................................................................................... 50
6.1 Characterization of spent NiMH batteries ................................................................................... 50
6.2 Recovery of metallic nickel .......................................................................................................... 50
6.3 Recovery of reusable AB5 alloy .................................................................................................... 50
6.4 Recovery of cobalt ....................................................................................................................... 51
7. FUTURE WORK ................................................................................................................................... 52
8. ACKNOWLEDGEMENTS....................................................................................................................... 53
9. BIBLIOGRAPHY .................................................................................................................................... 54
APPENDIX I: ANALYSIS TECHNIQUES ....................................................................................................... 60
I. A Qualitative analysis of crystalline compounds by X-ray diffraction (XRD) .................................... 60
I. B Scanning electron microscope (SEM) imaging of particles .......................................................... 60
I. C Elemental composition using inductively coupled plasma optical emission spectroscopy (ICP-OES)
........................................................................................................................................................... 60
I. D Solid sample surface area measurement using Brunauer-Emmett-Teller (BET) method ............ 60
I. E X-ray photoelectron spectroscopy (XPS) analysis of surface species ........................................... 60
vi
ABBREVIATIONS
Battery cycling continuous charge and discharge of battery cells
Ni(s) metallic nickel
AB5 alloy hydrogen storage alloy. In this work this refers to the AB5 type
Ni(OH)2 the cathode active material consisting of nickel hydroxide that has been
doped with Co and Zn. The particles of this material have been coated with
cobalt and this is also included in the abbreviation. Conductive nickel
network is not included in this abbreviation.
REE rare earth element. In this work REE refers to lanthanum, cerium,
praseodymium and neodymium
REE(OH)3 rare earth element (La, Ce, Pr, Nd) hydroxides
SHE standard hydrogen electrode potential
S/L solid to liquid ratio (g/L)
O:A ratio organic phase (volume) per aqueous phase (volume)
D distribution ratio
At% atomic percent
Leachate the produced aqueous solution containing elements that have been
dissolved from a solid sample
TBP tributyl phosphate
Cyanex 923 commercial blend of four tri-alkylphosphine oxides
ICP-OES inductively coupled plasma optical emission spectroscopy
XRD X-ray diffraction
EDS energy dispersive X-ray spectroscopy
SEM scanning electron microscopy
D2EHPA bis(2-ethylhexyl)phosphoric acid
TOA tri-octylamine
Cyanex 272 bis(2,4,4,-tri-methylpentyl)phosphinic acid
PC-88A 2-ethylhexyl 2-ethylhexyphosphonic acid
TBP tri-butyl phosphate
Alamine 336 tri-octyl/decylamine
Tween 80 surfactant
PAM polyacramide
vii
1. INTRODUCTION
The nickel metal hydride (NiMH) battery was commercialized at the end of the 1980s and has since then
been used in various applications [1-5]. Today, niche markets are energy back-up and smart grid
applications [6]. NiMH technology is a part of the present technologically diverse battery market that is
growing to meet the needs for adaptation to lower carbon footprint energy sources.
In the European Union (EU) the legislation demands general recycling of batteries in order to avoid
pollution by environmentally hazardous elements [7]. All types of batteries, from small button to large
industrial batteries, are collected, sorted and recycled.
Recycling of disposed NiMH batteries sustains the global markets with economic and technologic
important metals, i.e. nickel, cobalt and rare earth elements (REEs). These elements are used in various
applications, from common stainless steels to high tech alloys (i.e NiMH alloy) [8]. Examples of recycling
companies that treat discarded NiMH batteries are Nickelhütte Aue GmbH and Umicore [9, 10]. Both
pyro and hydrometallurgical methods are currently used today in order to produce pure metal fractions
from NiMH waste.
Sustainable production of metals and components through recycling is needed in order to decrease the
life cycle impact of these commodities. To increase sustainability of recycling it is necessary to selectively
recover metals and components, while minimizing the consumption of energy and chemicals.
NiMH components do not completely degrade throughout the battery lifetime and can potentially be
reused in new batteries. Reusing NiMH materials would add additional value to the waste, as producing
NiMH materials consumes high amounts of chemicals, energy and/or uses highly toxic compounds.
However, reusing spent NiMH electrode materials in new batteries has still not been practically
demonstrated.
The aim of the present work was to contribute to the development of selective and sustainable recycling
methods for NiMH electrode materials, thus making important parts of the electrodes reusable. The
first goal was to characterize the materials in respect to degradation of materials, in order to evaluate
the recycling option of reusing material. Further goals of the practical work were to investigate the
selective recovery of (1) metallic nickel, (2) reusable NiMH alloy and (3) cobalt.
2. BACKGROUND
2.1 Economic and environmental values of metals in nickel metal hydride (NiMH) batteries
Batteries are important products for the overall goal of making our energy consumption more
sustainable by storing energy from carbon-neutral energy sources. There are, however, economic and
environmental costs that are associated with the production of materials that are used in batteries. The
present work focuses on NiMH batteries and the emphasis on such material costs are therefore included
here.
Most NiMH battery components consist of inorganic materials. It is energy efficient to design the battery
as compact as possible. NiMH batteries therefore contain high concentrations of metals. As an example,
these batteries may contain 17.9 wt% nickel, 4.4 wt% cobalt and 17.3 wt% REE [11].
The most economically important metals used in NiMH batteries are nickel and cobalt. The metal
exchange market prices of 99.8-99.9 wt% purity nickel and cobalt are currently approximately 13000
and 33400 USD/tonne (respectively, in 2017) [12]. The energy consumption associated with the
production of a concentrate of nickel from sulfide (0.5-7 wt%Ni) and laterite (1-3 wt%Ni) ores are high
(60-1300 and 125-400 GJ/ton, respectively) [13]. The different energy consumptions are related to the
1
concentration of nickel in the ore. As can be expected, the subsequent refinement methods to produce
pure nickel (99 wt%) are much less energy demanding (10-38 GJ/ton).
Cobalt is a byproduct from primary production of copper and nickel ores, which are used for the
production of 35% and 55%, respectively, of global cobalt production [14].
Although REEs are not of immediate economic concern their general availability has been frequently
discussed in relation to the development of important technologies and growth of clean energy
economies [8, 14, 15]. Examples of applications where REEs are important are batteries and magnets.
Battery alloys, such as the ones used in NiMH batteries, have been reported as a significant market for
the consumption of REEs in the last decade [16, 17]. According to Binnemans et al. [16] global recycling
of REEs is low (< 1% in 2011). Recycling rates for REEs need to be improved as the primary production
of REEs (La, Ce, Pr, Nd) is an energy intensive process (687 GJ/ton) [18, 19].
An assessment by Majeau-Bettez et al. [20] showed that the production of NiMH batteries contribute
significantly more (> 55%) to the overall environmental impact of NiMH batteries compared to the use
phase. Unsurprisingly, the metal depletion potential was the highest score for the considered
environmental impacts.
Treating NiMH waste that has high concentrations of nickel and REEs can have a significant impact on
the life cycle energy consumption for these metals. Consequently, the environmental impacts of NiMH
components, which are rich in these metals, should also decrease. Managing NiMH waste also
recirculates economic and technologically important metals back into society.
2.2 Fundamentals of the NiMH battery

The fundamental components for a NiMH battery are the active anode (hydrogen storage AB5 alloy) and
cathode (Ni(OH)2) electrode materials, separator and electrolyte (KOH/LiOH). The NiMH battery is
discharged by the coupled anode (M=AB5) and cathode reactions,
MH + OH- ⇄ M + H2O + e- Eq. 1 (anode) E° = -0.83 V
NiOOH + H2O + e- ⇄ Ni(OH)2 + OH- Eq. 2 (cathode) E° = 0.52 V
MH + NiOOH ⇄ M + Ni(OH)2 Eq. 3 (total cell reaction) E° = 1.35 V
As can be seen in Eq. 1 and Eq. 2 water carries hydrogen ions between the electrodes. Hydrogen is
reduced during charge and oxidized during discharge which results in a zero net consumption of water
(Eq. 3). The reduced hydrogen does not combine into hydrogen gas but diffuses into the AB5 alloy until
released during discharge (Eq. 1).
There are also other reactions that occur inside a NiMH battery during practical operation that need to
be addressed when designing these batteries. In principle, as soon as any current is applied in the cell
several temperature, concentration, voltage and current gradients are created [21]. These variations
promote some competing reactions to occur depending on the charge/discharge conditions [21-23].
The reactions that are of concern during the charge and discharge processes are shown in Figure 1.
The cathode and anode interact with each other not only via the electrolyte but also through the gas
phase [22]. Gas pressure may increase, due to evolution of oxygen and hydrogen at the cathode and
anode (respectively) [24]. Excess anode capacity does not allow hydrogen gas to be evolved at the anode
during overcharge, because the cathode has already been completely charged. The excess anode also
allow recombination of oxygen and hydrogen gases at the anode surface; a reaction that produces
water. Using a starved electrolyte design allows some porosity in the separator and gases are able to
2
pass between the electrodes. As a precaution the closed cells are also equipped with a safety vent.
Although releasing the overpressure may be necessary, this quickly dries out the cell.
Figure 1. The principal reactions during charging and discharging of the sealed NiMH battery cell [24].
2.2.1 Hydrogen storage AB5 alloy used in the anode

There are many different formulations of the hydrogen storage alloy used in NiMH batteries. One of the
common types is the hexagonal AB5 with CaCu5 crystal structure (see Figure 2). The amount of A (e.g.
La, Ce, Pr, Nd) atoms are proportionally less than B (e.g. Ni, Co, Mn, Al) elements [25]. The type of
element and/or amount used for A and B can vary between producers.
The starting materials for the AB5 alloy are pure metal ingots, except for the ingot containing REEs (La,
Ce, Pr and Nd). A mixture of REEs is acceptable to use since these elements have similar chemical
properties. The similar chemical properties of REEs is due to the lanthanide contraction where the
atomic and ionic radii, respectively, are relatively constant from lanthanum to lutetium [26]. Avoiding
the need for elemental separation during primary production result in a much cheaper AB5 alloy.
Producing AB5 alloys involves the following steps: mixing of ingots, melting in inert atmosphere, rapid
cooling and finally reducing particle size by either mechanical crushing or hydrogen sorption/desorption
cycles [27]. In order to produce a material with strictly homogeneous composition it is also useful to
post-anneal the as-cast ingot before the particle size reduction [28]. This process has been reported to
proceed for 10 h at 1000 °C, according to Fetcenko et al. [25].
Clean surfaces of metals are needed for dissociative chemisorption and associative desorption of
hydrogen, according to Schlapbach et al. [29]. The necessary surface properties of the AB5 alloy particles
have also been discussed by Ye et al. [21]. The surface has to form a dense layer that inhibits corrosion
in the alkaline electrolyte during battery applications. However, it also needs to remain catalytically
active to create and break O-H bonds in the water molecules. Additionally, some hydrogen permeability
is needed in the AB5 bulk/surface interface as hydrogen atoms are moved between the interface during
battery cycling. Finally, it is also important that the surface is electrically conductive. As pointed out by
Ye et al. [21] there is little knowledge of the reactions that take place on the particle surfaces.
3
It is generally accepted that high-rate discharge of the negative electrode is influenced by the reaction
kinetics at the AB5 alloy surface [30]. A large reaction surface and a uniform particle size are important
for the high-rate performance [31]. These two properties lead to a minimization of various gradients in
the cell [21].
Producing the AB5 alloy material used for NiMH batteries from ingots is an energy intense process. It is
therefore of interest to reuse this material in new batteries.
Figure 2. The principal crystal structure of hydrogen storage alloy AB5 hexagonal P6/mmm. Constituents of A and B may vary in
elements and amounts, and between producers. Here, an example of formulation is A = La, Ce, Pr, Nd (blue) and B = Ni, Co, Al,
Mn (gray and green).
2.2.2 Nickel hydroxide Ni(OH)2 used in the cathode

The aim of this section is to review the properties of nickel hydroxide used for the NiMH cathode active
material.
Due to the low conductivity of nickel hydroxide efforts have been made to improve its conductivity. A
specific focus has been how the nickel hydroxide material behaves during cycling. The simplified
sequence of phase transitions (see Figure 3) was first described by Bode et al. [32]. Since then the
understanding of the complexity of nickel hydroxide has increased and a good review has been
published by Hall et al. [33].
charge
β-Ni(OH)2 discharge
β-NiOOH
dehydration overcharging
charge
α-Ni(OH)2 γ-NiOOH
discharge
Figure 3. Phase transitions of nickel hydroxide in alkaline media and battery applications [34]. According to Hall et al. [33] the α-
Ni(OH)2 is better described as α-Ni(OH)2⋅xH2O (0.41≥x≥0.7). Dehydration refers to the removal of intercalated water between the
Ni(OH)2 planes.
As can be seen in Figure 3, different phase transitions in each charge and discharge state are possible
[35]. The crystal parameters of these phases can be seen in Table 1. The crystal lattice parameter (c) for
α-Ni(OH)2 (8.0) and γ-NiOOH (6.9) are large compared to β-Ni(OH)2 (4.6) and β-NiOOH (4.8). Preferably,
the cathode electrode should only operate in β-Ni(OH)2/β-NiOOH cycling, because the lattice volume
change between this pair is lower compared to that of the α-Ni(OH)2/γ-NiOOH pair. According to Chen
et al. [36] the lattice volume increases by 44 vol% during the transition of β-NiOOH to γ-NiOOH.
4
The increased cathode volume due to cycling between the two oxidized/reduced forms of nickel
hydroxide can cause the electrolyte balance in the cell to shift, and electrolyte is redistributed toward
this electrode. As a consequence the internal resistance of the cell increases [37, 38]. According to
Shinyama et al. [39] this is one of the dominating factors for battery failure.
Table 1. Parameters for crystal structures of different phases of nickel hydroxide [40, 41].
Phase a = b (Å) c (Å)
β-Ni(OH)2 3.13 4.6
α-Ni(OH)2 3.02 8.0
β-NiOOH 2.81 4.84
γ-NiOOH 2.82 6.9
Partial substitution of nickel ions with zinc and cobalt ions is a common procedure to improve the nickel
hydroxide performance in batteries. In particular, it is believed that zinc suppresses formation of γ-
NiOOH, and that cobalt increases conductivity of the material by formation of CoOOH during the
charging process [36, 42, 43]. It has also been reported that co-precipitation of cobalt minimizes oxygen
evolution [44]. Oxygen evolution may cause oxidation of the metallic anode components, e.g. AB5 alloy.
To manufacture a spherical form of nickel hydroxide to be used in NiMH batteries it is necessary to

control the growth of particles during synthesis. A popular method is to feed a stirred reactor with nickel
salt solution and an alkaline chemical. The addition of dopants (Co and Zn) is easily achieved by adding
salts of these elements into the nickel solution. Complexing agents temporarily enhance solubility that
assist the Ostwald growth process to produce spherical particles [45, 46]. After filtration the collected
particles are dried and can be used without further treatment. A summary of suggested process
parameters used by various authors can be seen in Table 2. The resulting tapping density of the
produced particles refers to how efficiently the particles can be packed.
Table 2. A summary of process parameters for the synthesis of Ni(OH)2 particles used in NiMH batteries.
Temp. Chem. time Drying Size Tapping density

Metal salts (°C) pH Atm. Base additive (h) (°C) (µm) (g/cm3) Ref.
max 5M 10 M
SO42- 60 N2 10-20 n/a n/a 1.8 [47]
11.4 NaOH NH4OH
SO42-, NO3-, Cl- 50-601 111 n/a NaOH Not used n/a 1201 2-30 2.2 [48]
C2H3O2-, NO3-,
80 n/a Ar n/a NH4OH 3 80 n/a n/a [43]
Cl-, SO42-
4M
C2H3O2-, SO42- 50 9-13 N2 Tween 80 14 n/a 5-100 n/a [49]2
NaOH
4M
SO42- 50 n/a n/a PAM 14 120 27 2.3 [50, 51]2
NaOH
11-
SO42- 50-55 n/a NaOH NH4OH 60 105 n/a n/a [46]
11.5
SO42- n/a n/a n/a NaOH n/a n/a n/a 13 2.3 [52]
12-
SO42- 55 n/a NaOH NH4OH n/a 80 8 2.1 [53]
13
n/a not stated or not described in detail.
1stated as optimum parameters.
2non spherical particles were produced.
Cobalt is also applied as a coating on the nickel hydroxide particle surface in order to increase the
conductivity, which results in higher utilization of the electrode material. Utilization refers to the
amount of β-NiOOH that is discharged compared to the theoretical amount (289 mAh/g) [43]. Oshitani
et al. [54] first reported a method to coat the particles with cobalt, and since then other methods to
coat the particles have been suggested [55-58]. Most authors seem to be in agreement that the final
5
compound is CoOOH, although the reduction-oxidation reactions that would produce these surface
layers do not appear to be well understood. Compared to less conductive NiOOH (10 -1 mho/cm) the
CoOOH has higher conductivity (50 mho/cm) [59].
2.2.3 Metallic powders nickel and cobalt

Metallic nickel is used to increase the electron transfer between particles in both electrodes, as well as
to the current collector. The active materials are pressed into foam nickel substrates or mixed with
nickel fibers [24], i.e. Ni(s). These conductive networks are important for the discharge performance,
and nickel fibers can be used for ultra-high power discharge.
Cobalt metal powders can also be added to the cathode formulation. It has been reported that the
metallic cobalt re-precipitates as CoOOH during the first charge/discharge cycle, which then improves
the life-cycle stability of the cell [60].
There are several methods to produce cobalt powders. The common industrial processes are the Sheritt
process, solid phase reduction, the amalgam technique and electrolytic deposition [61, 62]. The
carbonyl process (Mond Process) can be used to produce cobalt powders. The reaction with cobalt and
carbon monoxide occurs at 150-170 °C [61]. This process is however commercially restricted for cobalt
due to low capacity and poor quality of the obtained powder [62]. In the Sheritt process a sulfide
concentrate is leached under pressure at 90 °C with ammonium sulfate. High pressure hydrogen is
injected and cobalt is recovered as a metallic precipitate.
In the production of nickel powders used in NiMH batteries the carbonyl process is the most important
commercial production method [62]. Nickel tetra carbonyl is formed by allowing carbon monoxide to
pass over metallic nickel at 50 °C [63]. Nickel tetra carbonyl is liquid at room temperature. The
decomposition into metallic nickel occurs in heated atmospheres at around 230 °C. By controlling the
process parameters it is possible to recover particles with various shapes (i.e. fibrous or spheres) and
size (less than a few microns). More information about the process can be found in the original research
paper by Mond et al. [64] and other references [65, 66].
2.3 Degradation mechanisms of battery cells and materials

NiMH batteries are eventually discarded when the required standard for application is no longer met.
The overall degradation of the cell is a complex process that involves different phenomena, e.g. loss of
electrolyte, increased internal resistance, corrosion of the AB5 alloy, pulverization and contamination
between the electrodes. There are numerous scientific papers that discuss NiMH battery degradation
[21-23, 30, 31, 37-39, 67-78].
To simplify this review, the main degradation aspects of the electrode materials are separately
considered in the following sections. However, it should be emphasized that battery degradation
includes all the separate processes described above. For this reason references will reoccur in the
following subsections.
6
2.3.1 Loss of electrolyte
One of the main causes of NiMH battery degradation is that the cell dries out, i.e. the electrolyte liquid
is lost. The purpose of the electrolyte is to provide ionic conductivity and the optimal concentration is
approximately 6 M KOH [68]. Above this concentration the discharge capacity of the anode electrode
decreases.
During charge and discharge of the battery the electrolyte redistributes between the electrodes, causing
the separator to dry out. As a consequence, the internal resistance of the cell increases [37, 38]. As
previously mentioned, increased internal resistance is believed to be one of the dominating factors for
battery failure [39]. In agreement with this notion Li et al. [38] replenished the electrolyte of spent
batteries and demonstrated that the discharge capacity could be partially restored to 10% of the original
values. This demonstration indicates that there are other causes for battery failure. However, it is
reasonable that a lack of electrolyte plays an important role by promoting unwanted reactions in the
electrodes, due to increased temperature caused by increased internal resistance.
2.3.2 Pulverization
Pulverization refers to particle size reduction through mechanical (i.e. grinding) or chemical treatment.
The effect of chemical treatment is that the lattice volume contracts or expands unevenly, which results
in parts of the particle bulk cracking. An example of chemical treatment that results in pulverization is
hydrogen intercalation into lattices of strong hydride forming elements (i.e La, Ce, Pr, Nd) [79].
Both cathode and anode materials are pulverized during continuous charge/discharge of the battery. It
is the phase transitions that occurs during cycling that are the main cause of this degradation process.
Pulverization of the cathode electrode material increases with increasing number of cycles [38, 69]. Li
et al. [38] noticed an increase of γ-NiOOH phase formation in the cathode electrode during cycling. As
mentioned previously, this phase is associated with high (44%) lattice volume change.
Pulverization of the AB5 alloy has been observed by many authors [23, 30, 38, 69, 71]. According to
Bäuerlein et al. [23] both the chemical composition of the alloy and the amount of intercalated hydrogen
affects the increase of lattice volume during charge. Pulverization is caused by mechanical stress within
the alloy, which builds up during cycling. As a final result, the alloy particles crack. Since consecutive
cycling reduces the AB5 alloy particle size, the AB5 alloy surface area also increases. Both preferable
electrochemical and corrosion reactions are promoted as a consequence of these physical changes.
Interestingly, it has been demonstrated that a decreased AB5 alloy particle size increases the discharge
capacity of the anode electrode [23]. Although the cycle stability of the anode is also decreased, spent
AB5 alloy can potentially be reused for high-discharge rate applications.
2.3.3 Corrosion
Corrosion of the AB5 alloy material has been in focus for research into the causes of battery failure.
Authors [23, 37, 74] generally illustrate the corrosion of the AB5 alloy as being similar to the following
reaction (M = AB5),
2𝑀 + 𝐻2 𝑂 → 𝑀𝐻 + 𝑀𝑂𝐻 Eq. 4
It was suggested that the initial corrosion is enhanced by the lack of protective corrosion products on
the alloy particles [37]. Interestingly, it has also been reported that pre-treating the alloy in alkali
solution at elevated temperature improves the life-cycle of the AB5 alloy by the formation of protective
corrosion layers [71]. These reports suggest that formation of metal oxide/hydroxide layers are
important for inhibiting further corrosion. However, the thickness and composition of these layers can
potentially retard e.g. hydrogen diffusion rates [23].
7
The reported compositions of the corrosion layers that form on the alloy particles during cycling appear
to be similar [37, 69, 71, 72]. Zhou et al. [69] detected three phases with different degrees of oxidation;
(1) low oxidized core of AB5 particles, (2) partially oxidized with very little manganese and aluminum
content and (3) heavily oxidized with mostly REE and manganese. Ikoma et al. [71] observed that
REE(OH)3 and manganese oxides were formed at the alloy surface and that nickel and cobalt existed in
metallic form near the surface. It was proposed that manganese initially dissolved and then re-deposited
as an oxide on the alloy surface. Similar observations of corrosion products were made by others [23,
72]. Furthermore, Maurel et al. [72] reported that grains of MnO(OH) (s), Mn(OH)2 (s) and CoOOH (s)
were randomly scattered over the alloy surface and covered the REE(OH)3 needles. Based on the analysis
work of Young et al. [78] it can be approximated that the thickness of the surface layers are in the order
of 50-100 nm. According to Schlapbach et al. [29] it is corrosion products like the ones mentioned above
that inhibit the critical function of the AB5 alloy to absorb and desorb hydrogen.
Removing corrosion products is important for the possibility to reuse the AB5 alloy, in order to allow
sufficient hydrogen diffusion rate to and from the AB5 bulk.
2.3.4 Cross contamination between cathode and anode

Contamination of metals from one electrode onto the other has been reported to occur during battery
cycling.
Leblanc et al. [37] suggested that measurements of aluminum in the cathode can be used as an indicator
of corrosion of the AB5 alloy. Since that report was published this method has been adopted by others
[23, 72]. Of course, this method of quantifying AB5 corrosion is only valid if all the oxidized aluminum
can be found in the positive electrode. Some authors state that as aluminum is dissolved it will end up
trapped in the Ni(OH)2 lattice [37, 69, 74, 75]. Aluminum is an unwanted contaminant that is known to
stabilize the unwanted α-Ni(OH)2 structure during cycling [69].
Contamination of manganese in the cathode electrode has also been observed [23]. According to the
investigation performed by Gerlou-Demourgues et al. [80] substitution of nickel with manganese can
also lead to the formation of α-Ni(OH)2.
The anode is also susceptible to contamination of elements from the cathode. Zhou et al. [69] reported
contamination from the cathode, through observations of zinc on the spent AB5 alloy material. This
element does not appear to have a detrimental effect on the anode electrode. Instead, it may even
inhibit further corrosion [74].
8
2.4 Hydrometallurgical recycling methods to treat NiMH waste
2.4.1 Dissolution
Dissolution of spent NiMH waste is a method to produce an aqueous solution (i.e. a leachate), which
will be subsequently treated using separation and recovery methods. It is also the first opportunity for
chemical separation. Selective separation is economically interesting because it reduces the need for
subsequent treatment.
Interestingly, only mineral acids (hydrochloric, sulfuric and nitric acid) have been considered for
complete dissolution of the waste, in previous investigations (see Table 3).
Table 3. Optimal leaching conditions with respect to the dissolution of nickel and cobalt. Mixed refers to mixed cathode and
anode material. For the sake of comparison values have been adapted from references through both tables and graphs.
Differences in reported REEs leaching results are due to differences in consideration of the reaction yields (i.e. dissolution or
oxidation).
Electrode(s) Acid Temp. °C time S/L (g/mL) Ni Co REE Ref.
1.5 M H2SO4 + 4 vol%
Mixed 50 1h 1/50 100% n/a n/a [81]
H2O2
Anode, cathode 2 M H2SO4 20 2h 1/10 76.6% 97.6% > 90% [82]
Mixed 3 M H2SO4 75 5h 1/10 > 95% >95% > 95% [83]
3 M H2SO4
Mixed 95 4h 13/100 > 95% >95% < 5% [52]
Mixed 2 M H2SO4 95 4h 5/100 97% 100% 96% [84]
Mixed 3 M HCl 95 3h 1/10 97.4% 100% 99.1% [85]
Mixed 1.5 M H2SO4 30 1h 1/20 97% 95-100% 87% [86]
Anode 3 M H2SO4 95 4h 1/10 99.5% 99.5% 94.5% [87]
Mixed 2 M H2SO4 20 2h 1/10 > 76.6% >97.6% > 91.8% [88]
Mixed 12 M HCl 40 2h 15/100 n/a n/a n/a [89]
Mixed 4 M HCl 95 3h 1/10 98% 100% 99% [90]
Mixed 4 M H2SO4 80 3h n/a 100% 98% 36% [91]
Anode, cathode,
pH ≥1: H2SO4, HCl, HNO3 30 5.5 1/50 n/a n/a n/a [92]
mixed
Mixed 8 M HCl 30 n/a 3.7/10 n/a 100% 100% [93]
n/a – not available or not presented in detail.
Some general observations can be from the work presented in Table 3. Firstly, mineral acids
(hydrochloric, sulfuric and nitric acid) have been extensively studied and it is expected that these acids
will efficiently dissolve the mixed electrode materials. Secondly, it is possible to separate REE sulfates
from sulfate solutions [52]. Thirdly, the S/L ratio can generally be kept high (1-3.7/10) in both
hydrochloric and sulfuric acid solutions. A high S/L is preferable, as the dissolution step produces less
9
secondary liquid waste that needs to be treated. No other acidic media than the above mentioned ones
have been found to have been extensively studied.
It is worth further mentioning the work by Larsson et al. [92] because these authors investigated in
detail the dissolution behavior of electrode materials in hydrochloric, sulfuric and nitric acid. At high pH
(≥ 1), treatment in these acids will result in different outcomes. In nitric acid all material is dissolved
after 14 hours at pH 1. Dissolution of metallic nickel is expected because nitric acid is an oxidizing acid.
In de-aerated hydrochloric and sulfuric solutions, the dissolution of metallic nickel is suppressed.
Interestingly, de-aereated conditions do not seem to be necessary in practice. The required time for
dissolution of the waste matrix was approximately 5.5 h in aerated solutions, using either sulfuric or
hydrochloric acid at pH 1 [92]. In the work by Larsson et al. [92] the metallic nickel was of the foam type,
and its dissolution did not appear to be complete even after 44 h in hydrochloric acid. Therefore, it may
be possible to practically separate the Ni(s) component even in aerated solutions.
2.4.2 NiMH metal separation by solvent extraction

Solvent extraction techniques have been used to treat leachate solutions that have been produced by
dissolution of NiMH waste into acidic solutions. In a solvent extraction separation of elements dissolved
in an aqueous leachate, an immiscible organic phase is brought into contact with the leachate solution.
The organic phase contains an extracting agent, which is able to form a complex with the metal ion that
is more soluble in the organic phase than in the aqueous phase. Thus, the metal ion is extracted, i.e.
transported to the organic phase.
Several extracting agents have been proposed for the extraction and separation of elements from
leachate solutions [52, 81, 84-87, 89, 90, 93-98]. Some of these suggested systems have been optimized
for practical recycling applications. A summary of some of these systems are given in Table 4. As can be
seen from this table there are several extraction agents that can be used to separate nickel, cobalt and
REEs from sulfate and chloride aqueous solutions.
Table 4. Recovery rates (%) of rare earth elements (REE), cobalt and nickel according to suggested extraction systems.
Leachate anion REE (%) Cobalt (%) Nickel (%) Extracting agent system Ref.
93,6 96 96 D2EHPA, Cyanex 272 [84]
97,8 98,1 98,1 D2EHPA, Cyanex 272 [52]
SO42-
98,9 98,5 98,4 PC-88A [87]
99,5 99,5 99,6 Quaternary ammonium salts [97]
98 98 96 D2EHPA, TOA [85]
Cl- 98,9 93,6 97 TBP, Alamine 336, PC-88A [89]
99,9 99 99 Cyanex 923 [93]
Recovery of metals from a loaded solvent can be made by stripping (i.e. back-extraction) [99]. For
stripping it is important to make extracted metals soluble in the aqueous phase, and approaches are
dependent on the system at hand. Examples of methods to strip metals are to vary the pH, amount of
aqueous phase, temperature and/or use counter-ions (e.g. nitrates). Stripping of metals can result in
elemental separation and therefore the stripping step is appropriately applied before recovering the
metals as solid compounds.
10
2.4.3 Recovery of metals
Recovery of metals from leachate solutions produced through hydrometallurgical treatment of NiMH
waste has been investigated using different techniques. These methods include precipitation,
calcination and electrowinning of metals. Electrowinning refers to the extraction of metals from
compounds through the electrolysis of solutions containing metal ions [100]. Applying a current that
matches the reduction potential of a metal ion can reduce the metal ion into its elemental state.
Tzanetakis et al. [90, 95] investigated electrowinning of nickel and cobalt from aqueous chloride media
at pH 3. Depending on the current efficiency the composition of the nickel (76-83%) and cobalt (11-
14%) varied. Lupi et al. [101] also reduced nickel and cobalt, but from sulphate solutions at pH 4.3.
Recovery of metals can be made by adding precipitation agents into the leachate solutions. Metal ions
form insoluble complexes and precipitate out of solution. Once nickel, cobalt and REE have been
separated (using i.e. solvent extraction separation) these purified metal fractions may be recovered
using oxalic acid [84, 85, 89, 97]. To transform these solids into oxides, the recovered material can be
calcined at 850-900 °C [84, 85]. Nayl [97] reported that the calcination of oxalates of REE, cobalt and
nickel can be performed also at lower temperatures (700, 500 and 400 °C, respectively). Precipitation
using alkaline solutions has also been discussed by others [52, 83, 86, 91, 94, 96, 102-104].
3. THEORY
3.1 Oxidation of metals
The current work discusses the recovery of Ni(s) and AB5 alloy using acidic media. It is difficult to recover
metallic NiMH components due to the generally low reduction potentials, which are shown in Table 5.
Consider the half cell reactions of the least reactive system of Ni(s) and water,
𝑁𝑖(𝑠) ⇄ 𝑁𝑖 2+ (𝑎𝑞) + 2𝑒 − Eq. 5
2𝐻 + (𝑎𝑞) + 2𝑒 − ⇄ 𝐻2 (𝑔) Eq. 6
Here, the reduction of hydrogen (Eq. 6) represents the splitting of water and is not associated with acid
consumption. The Nernst equation can be used to calculate the cell potential of Eq. 5 and Eq. 6 [105],
𝑅𝑇
𝐸 = 𝐸0 + 2.3 𝑛𝐹 𝑙𝑜𝑔(𝑄 −1 ) Eq. 7
where the constants R, T and F represents the gas constant (8.3141 J mol -1 K-1), temperature (K) and
Faradays constant (96.485 J mol-1 V-1), respectively. The variable n is the number of electrons transferred
per mole reaction. E0 is the standard reduction potential. The reaction quotient Q is defined as,
𝑐 𝑑
𝑎𝐶 𝑎𝐷
𝑄= 𝑎 𝑎𝑏 𝑤ℎ𝑒𝑛 𝑎𝐴 + 𝑏𝐵 ⇄ 𝑐𝐶 + 𝑑𝐷 Eq. 8
𝑎𝐴 𝐵
For an ideal gas, ideal solution or pure solid/liquid, the activity (a) can approximated by its partial
pressure (P/P0, P0 = 1 bar), molar concentration (M/M0, M0 = 1 mol/L) or by unity, respectively [106].
The hydrogen partial pressure is per definition at unity, in order to compare cell potentials with the
standard hydrogen electrode (SHE).
To determine oxidation of metals it is conventional to choose metal concentrations of 10-6 M [105]. The
cell potential of nickel and water in the least acidic system available (pH ≈ 7) can be calculated,
𝑅𝑇
𝐸 = 𝐸𝐻 + − 𝐸𝑁𝑖2+ + 2.3 (log(𝑎𝑁𝑖(𝑠) ) + 2log(𝑎𝐻 + ) − log(𝑎𝑁𝑖2+ ) − log(𝑎𝐻2 )) = [Table 5] =
𝑛𝐹
11
0.059 𝑉
= +0 𝑉 − (−0.26) 𝑉 + (log(1) − 2𝑝𝐻(= 7) − log(10−6 ) − log(1)) =
2
= 0 𝑉 + 0.26 𝑉 − 0.236 𝑉 = 0.024 𝑉
The cell potential is positive which means that nickel will oxidize in any aqueous media at pH ≤ 7.
The AB5 alloy contains elements (Co, Al, Mn, REEs) with reduction potentials below that of nickel.
Therefore it is reasonable to also assume that the AB5 alloy component will also oxidize in acidic media.
It can be concluded that the recovery of AB5 alloy and Ni(s) using acidic media relies completely on the
difference in dissolution rates of the NiMH battery components.
Table 5. Standard reduction potentials for elements found in NiMH batteries vs the standard hydrogen electrode [105, 107].
Oxidized ⇄ Reduced E° (V) vs SHE
Li + e ⇄Li
+ - -3.04
K + + e- ⇄ K -2.93
La3+ + 3e- ⇄ La -2.38
Pr3+ + 3e- ⇄ Pr -2.35
Nd3+ + 3e- ⇄ Nd -2.32
Ce3+ + 3e- ⇄ Ce -1.72
Al3+ + 3e- ⇄ Al -1.66
Mn2+ + 2e- ⇄ Mn -1.19
Zn2+ + 2e- ⇄ Zn -0.76
Co2+ + 2e- ⇄ Co -0.28
Ni2+ + 2e- ⇄ Ni -0.26
2H+ + 2e- ⇄ H2 -0.059pH [105]1
O2 + 4H+ + 4e- ⇄ H2O +1.23
1Not to be confused with the hydrogen reduction associated to acid consumption.
3.2 Recovery of nickel

3.2.1 Recovery of metallic nickel using hydrochloric solutions
Recovery of Ni(s), i.e. the conductive nickel particle network used in the electrodes, from NiMH waste
using acidic solutions can be achieved if competing surface processes limit the oxidation of Ni(s). In the
system considered here, the most important process that can protect the metal surface is adsorption
of hydrogen.
Nickel is expected to dissolve into acidic media, e.g. hydrochloric acid, through two general reactions,
𝑁𝑖(𝑂𝐻)2 + 2𝐻𝐶𝑙 (𝑎𝑞) → 𝑁𝑖𝐶𝑙2 (𝑎𝑞) + 2𝐻2 𝑂(𝑙) Eq. 9
𝑁𝑖(𝑠) + 2𝐻𝐶𝑙 (𝑎𝑞) → 𝑁𝑖𝐶𝑙2 (𝑎𝑞) + 𝐻2 (𝑔) Eq. 10
These reactions (Eq. 9-10) are also expected to occur for other metals found in the NiMH waste.
Hydrogen adsorbs strongly only to nickel surfaces, and the nickel hydride bond can be approximated to
-96 kJ/mol [108, 109]. The strong nickel-hydride bond constitutes chemisorption (≤ -40 kJ/mol), which
is a binding strength that is characterized by less reversibility compared to physisorption (≥ -40 kJ/mol)
according to McCafferty [110]. That is to say, the hydride bond will block further acidic attacks. As
thermodynamics do not consider the kinetics of processes, one can only speculate that strong
chemisorption will results in a slow hydrogen desorption rate from metallic nickel surfaces.
The Ni(s) used in the current work is in the form of powder, and therefore has a high specific surface
area (m2/g). A high surface area can potentially compensate slow desorption rates of hydrogen, which
12
can result in a dissolution of Ni(s). Therefore it is not known if the nickel-hydride bond is sufficiently
strong to allow Ni(s) powder to be extracted using acidic media.
From practical considerations, hydrochloric acid is preferred over other conventional acids (sulfuric and
nitric acid). The practical properties of conventional acids can be seen Table 6. The NiMH waste consists
of predominantly nickel in the form of Ni(s), AB5 alloy and nickel hydroxide. In order to reduce leachate
volumes it is necessary that the nickel solubility is high in the leaching solution.
Between the different acids in Table 6, hydrochloric acid has the second highest solubility of nickel. The
chloride (-I) anion cannot be further reduced due to a completed outer electron configuration shell
(3p6). Since no reduction of the chloride can occur, the nickel hydride cannot be oxidized. However,
sulfates and nitrates do have reduction potentials above SHE, which suggests that these acids can
oxidize the nickel hydride. Therefore hydrochloric acid is preferred over sulfuric and nitric acid.
Table 6. Solubility of nickel salts produced by dissolution of nickel compounds into mineral acids [111].
Mineral acid Salt Solubility (water, 20 °C) [g/100 gwater] Ligand reduction [107]
H2SO4 NiSO4 40.1 𝑆𝑂4 2− / 𝑆𝑂3 2− +0.17 V
HCl NiCl2 60.8 None
HNO3 Ni(NO3)2 94.2 𝑁𝑂3 − / 𝑁𝑂2 − +0.93 V
𝑁𝑂3 − / 𝑁𝑂 +0.96 V
𝑁𝑂3 − / 𝑁2 𝑂4 +0.80 V
3.2.2 Distribution into the organic phase in solvent extraction

Separation of elements and compounds using solvent extraction is a technique that is used in
pharmaceutical and biochemical industries, analytical chemistry, metallurgy and industrial recycling of
inorganic waste [112].
By contacting a non-miscible organic phase with an aqueous phase (i.e. leachate), neutral complexes
distribute between the phases. A schematic representation of the solvent extraction technique is
illustrated in Figure 4. In the current work the extraction agent Cyanex 923 is used, and its extraction of
metals found in NiMH batteries is discussed in the following section.
Figure 4. Illustration of the solvent extraction technique. One metal (blue) is preferentially distributed in the organic phase over
another (red), after physical mixing and phase separation. The difference in distribution results in elemental separation.
The distribution of a metal A can be quantified using the distribution ratio at equilibrium,
[𝐴]𝑜𝑟𝑔
𝐷𝐴 = [𝐴]𝑎𝑞
Eq. 11
13
If 𝐷𝐴 >> 1 the metal is mainly distributed in the organic phase. Separation of metals from a mixed
leachate solution occurs when the distribution ratios of the metals are different.
3.2.3 Extraction of chloride complexes with Cyanex 923

Nickel is the most abundant metal in NiMH batteries and extracting agents are relatively expensive. It is
therefore economically beneficial to extract all other elements except nickel from leachate solutions.
The commercial mixture of tri-alkylpshosphine oxides sold under the name Cyanex 923 can be used to
extract, for example, lanthanides and zinc from aqueous phases. This extracting agent is especially
useful for chloride based leachates.
The chloride-Cyanex 923 extraction system for separation of metals from NiMH waste was investigated
by Larsson et al. [113] at great lengths. In a first step zinc was extracted and separated from the other
metals by low (8 vol%) concentration of Cyanex 923. Extraction of remaining metals other than nickel
(Al, REE, Co and Mn) was then made by increasing the extractant concentration to 70 vol%. Recovery of
metals in groups from the organic phase was possible by contacting the organic phase with different
aqueous phases (i.e. stripping). In the first strip step cobalt and manganese were recovered with a
nitrate solution (0.1 M Na(NO3)2). The other elements (Al and REE) were subsequently recovered with
hydrochloric acid (1 M).
The extraction of metal chlorides with Cyanex 923 (𝐵) follows the solvating mechanism [113],
𝑀 𝑥+ (𝑎𝑞) + 𝑥𝐶𝑙 − (𝑎𝑞) + 𝑦𝐵(𝑜𝑟𝑔) ⇄ 𝑀𝐶𝑙𝑥 𝐵𝑦 (𝑜𝑟𝑔) Eq. 11
The aqueous metal chloride complex is associated with water, and it is more accurately described as
MClx(H2O)z. The water molecules are not co-extracted. Importantly, the extraction of i.e. lanthanum is
enhanced by increased chloride ion concentrations in the solution [93]. A complete neutralization of the
H+-ions from the acid in the dissolution of NiMH waste in 8 M HCl solution will create a chloride-rich
solution and thus promote extraction of the REEs.
In media that is too acidic the extraction of REEs may be low. According to Alguacil et al. [114] acid can
also be extracted by Cyanex 923 and that reduces the amount of ligand that is available for binding to
REEs,
𝐻 + (𝑎𝑞) + 𝐶𝑙 − (𝑎𝑞) + 𝐵(𝑜𝑟𝑔) ⇄ 𝐻𝐶𝑙𝐵(𝑜𝑟𝑔) Eq. 12
The practical implication of Eq. 12 is that stripping of aluminum and REEs will be affected by the co-
extracted acid. Any aqueous solution used for stripping metals (i.e. Co and Mn) from the loaded organic
phase will also strip the acid. Consequently, aluminum and REEs are also stripped from the organic
phase. For this reason alone, controlling the addition of acid in the prior dissolution step is important in
order to avoid excess acid remaining in the leachate.
3.3 Recovery of reusable anode using carboxylic acids

Carboxylic acids were chosen for this work to investigate the recovery of reusable anode (containing
AB5 alloy) material from NiMH waste. Here the intention was to dissolve other components, such as
nickel hydroxide and REE(OH)3, but leave the AB5 alloy relatively undissolved.
Trial experiments indicated that the spent AB5 alloy component was not immediately oxidized in
carboxylic acid solutions, which otherwise occurred in mineral acids. Smith et al. [115] suggested that
the AB5 alloy can be recovered selectively from a mixture of AB5 alloy and nickel hydroxide, by selective
dissolution of nickel hydroxide in carboxylic acid(s). In the work for the current thesis a demonstration
of the dissolution of NiMH waste in carboxylic acids has not yet been found.
14
Carboxylic acids have the functional group COOH, which results in the ability of these acids to
deprotonate in aqueous solutions and form metal complexes. Deprotonation of carboxylic acids
depends on the pH of the solution,
R-(COOH)x + xH2O = R-(COO -)x + xH3O+ Eq. 13
Deprotonation is suppressed at low pH values, which relates to the weak acid character of carboxylic
acid. As a practical example, in a 1 M carboxylic acid solution at pH 1 only 10 mol% of the acid is
deprotonated. This means that the acid solution will contain both COOH and COO- functional groups.
The stability constants for some carboxylic acids and metals from a critical survey by Martell et al. [116]
are presented in Table 7. This table shows the stability constants for deprotonation of some carboxylic
acids and their complex formation for nickel and lanthanum (major waste constituents). The fact that
stability constants have been quantified suggests that nickel and lanthanum carboxylates are soluble to
some degree.
Ultimately, the dissolution rates cannot be approximated by stability constants and need to be
investigated experimentally.
It is expected that dissolution of NiMH components in carboxylic acids is slow. Lone pair electrons on
the functional COOH group can interact with metallic surfaces through adsorption [110]. Inhibition of
carboxylic acids on metallic surfaces in mineral acids has been previously demonstrated using e.g.
malonic, ascorbic, citric and succinic acid [117-120]. The ability to donate electrons is an important
parameter for inhibition, and the greater the pKa the greater tendency for electron donation [110]. The
pKa of the carboxylic acids used in the current work can be seen in Table 7.
The effect of adsorption processes can result in a slow dissolution. Potentially, the adsorption onto
metallic surfaces, i.e. AB5 alloy, is stronger than onto hydroxide surfaces, thus resulting in a slower
dissolution of the AB5 alloy.
Table 7. Equilibrium quotients (e.g. stability constants) for the formation of metal complexes in different carboxylic acids. Values
are adapted from Martell et al. [116] for the range of 20-25 °C and 0.1-0.5 M.
Carboxylic acid pKa (1,2,3) Formula Equilibrium quotient Log K (Ni2+) Log K (La3+)
Acetic 4.6 C2H4O2 K = [ML]/[M][L] 0.7 1.8
K = [ML2]/[M][L]2 2.8
K = [ML3]/[M][L]3 3.5
Citric 2.9, 4.4, 5.7 C6H8O7 K = [ML]/[M][L] 5.4 7.6
3.3 [MHL]/[M][HL] 10.2 [ML2]/[M][L]2
1.8 [MH2L]/[M][H2L] 2.2 [MHL2]/[ML][HL]
Malonic 2.7, 5.3 C3H4O4 K = [ML]/[M][L] 3.2 3.7
K = [MHL]/[M][HL] 1.0
K = [ML2]/[M][L]2 4.9 5.9
Maleic 1.8, 5.8 C4H4O4 K = [ML]/[M][L] 2.0 3.5
K = [ML2]/[M][L]2 5.4
15
3.4 Recovery of cobalt using ascorbic acid
Cathode materials used in NiMH batteries are surface engineered with a conductive cobalt enriched
surface layer. It is believed that this layer still exists on (spent) cathode surfaces after battery cycling,
which consists of cobalt in oxidation states higher than 2. Ascorbic acid has both reducing and acidic
properties, and was chosen for the investigation of removing the cobalt layers selectively. It is expected
that the acidic property will react also with other battery components. To achieve selective dissolution
of cobalt it is therefore important that the simultaneous reduction and dissolution of cobalt is faster
than the dissolution of other components.
A practical challenge of selective dissolution in acids is the general dissolution reactions of both
electrode materials (Eq. 9-10). Since those reactions involve oxidation of the metals, it appears more
likely that a selective dissolution of cobalt-enriched surface layers will happen when using an acidic
reducing agent, namely ascorbic acid, than when using a common mineral acid.
Ascorbic acid can both dissociate its proton and become oxidized. According to Du et al. [121] the
dissociation of hydrogen ions and oxidization to dehydroascorbic acid are coupled, which can be seen
in Figure 5.
Figure 5. The oxidation of ascorbic acid to dehydroascorbic acid is coupled to hydrogen dissociation in aqueous solutions. The
image has been adapted and reconstructed from Du et al. [121].
If no oxidation of ascorbic acid occurs the formation of ascorbate ions is only dependent on the pH of
the solution. Speciation modelling of ascorbic acid in aqueous solution show that at pH 7 the dominant
species for ascorbic acid is AscH- (99.9%) and there are low concentrations of the other forms AscH2
(0.1%) and Asc2- (0.005%) [121].
The reduction potential of ascorbic acid has been reported to be in the range of –0.283 and -0.066 V, in
the corresponding pH range of 2-7 [122]. These reducing potentials do not raise any suspicion that
metals associated with NiMH batteries will be reduced from metal ions to elemental state in ascorbic
acid solutions. Common oxidation states of the metals used in NiMH batteries (see Table 5) does not
allow a significantly large positive cell potential to be formed between the metal ion and ascorbic acid.
16
The rate of redox reaction between ascorbic acid and cobalt is important for selective dissolution.
Ascorbic acid has not been tested previously for recycling NiMH materials. The ascorbic acid oxidizing
rate is therefore evaluated by broadening the discussion.
Buettner et al. [123] found that aerated and room-temperature aqueous solutions of ascorbic acid in
50 mM phosphate buffers (pH 7) experienced up to 30% loss of an original 125 µM ascorbate within 15
min. When care was taken to remove transition metals the loss of ascorbate was decreased to 0.05%/15
min. This reaction was not described in detail, but it is indicated that the oxidation rate of ascorbic acid
in the presence of transition metals is fast. In addition, the ascorbate ion can also oxidize by reaction
with soluble oxygen to produce superoxide through the following reaction,
𝐶6 𝐻6 𝑂6 2− + 𝑂2 → 𝐶6 𝐻6 𝑂6 ∙− + 𝑂2 ∙− Eq. 14
The observed pseudo-first-order rate constants of autooxidation of AscH- is k = 10-6 M-1s-1 and that of
Asc2- is k = 102 M-1s-1 [121]. This mean that the oxidation rate of ascorbic acid in the presence of oxygen
is most dependent on the presence of Asc2- ions.
There are reasons to suspect that simultaneous reduction and dissolution of solid cobalt (iii) into
solution should also be fast. According to Li et al. [124] the dissolution of cobalt from mixed lithium
cobalt oxide in ascorbic acid solutions is completed after 20 minutes of dissolution at room temperature
(20 °C).
17
4. EXPERIMENTAL
4.1 Experimental outline
This work focuses on characterization of the NiMH electrode materials and the feasibility of recovering
three different components, and these goals are listed in the following order:
(1) Characterization of unspent and spent cathode and anode.
(2) Recovery of Ni(s) from the conductive nickel network.
(3) Recovery of reusable anode material. Specifically, this is a mixture of AB5 alloy and Ni(s).
(4) Selective recovery of cobalt from cobalt-enriched surface layers on cathode materials.
These recovery processes involve the use of different chemical systems, which are discussed separately.
A schematic overview of the experimental work is shown in Figure 6.
The materials used for the manufacturing of the cathode electrode are powders of Ni(s), metallic cobalt
and nickel hydroxide. The anode is made only from Ni(s) and AB5 alloy. The respective electrode
mixtures had been pressed onto polymeric carrier sheets.
Batteries were charged and discharged (i.e. cycled) by the battery producer Nilar AB [6]. The batteries
were considered spent when the internal resistance was twice that of the initial resistance, below 80%
of capacity and/or due to miscellaneous aspects causing battery failure.
Unspent materials were assembled in the same process as the cycled batteries. However, unspent
materials were not subjected to electrolyte (25-30% KOH).
A total of 5 different types of samples were produced, namely unspent cathode and anode, spent anode
and cathode, and spent mixed electrode materials. One spent cathode and one spent anode batch were
produced by collecting the respective separated material from at least 3 batteries. These batches were
used to produce mixed electrode samples.
18
Figure 6. Process flow diagram for experimental work and samples presented in this thesis. Unspent materials originate from
batteries that have been assembled and disassembled. Spent materials refers to batteries with charge and discharge properties
that are considered below standards for battery applications.
4.2 Battery disassembly

Discharged bipolar NiMH batteries of the size 12 V and 10 Ah were provided by Nilar AB. A schematic
overview of this type of module is shown in Figure 7. The modules had been cycled differently for various
application purposes.
Prior to disassembly the spent batteries were discharged in the laboratory using a 50 Ω resistor until
less than 2 V and 3 mA could be measured using a conventional multimeter. The active materials were
manually separated from the casing, separator, terminal and current collector into separate anode and
cathode fractions. Depending on the use phase of the battery the disassembled anode fractions can
react with oxygen in normal atmosphere and self-ignite, and therefore this fraction was immediately
immersed into water (>18 MΩ/cm, Milli-Q, Millipore).
Water was used to rinse the fractions in order to remove the remaining KOH-rich electrolyte. During
this procedure the electrode substrate (polymeric carrier sheets) was removed. The produced materials
were dried at 55 °C and normal atmosphere until no further weight loss could be detected.
A reference battery with no addition of electrolyte was built, disassembled and used without further
work up procedures.
19
Figure 7. Schematic overview of NiMH battery module illustrating the bipolar stack design [6]. The illustrated components (left
to right) are casing, terminal, current collector, anode, separator, cathode, separator, inner and outer stack frame.
4.3 Characterization
Characterization of spent cathode and anode materials was done to investigate the chemical and
physical state of the spent materials in order to develop specific treatment methods. In addition,
characterization of unspent and spent materials was necessary to identify degradation, in order to
evaluate the feasibility of recovery and reuse of the AB5 alloy. Several characterization techniques were
used to investigate degradation of the electrode material and are listed in Table 8.
Table 8. Investigated chemical and physical property with respective analysis technique. Four types of materials were
characterized separately: unspent cathode, spent cathode, unspent anode and spent anode. Analysis techniques are described
in Appendix I.
Property/ Degradation aspect Analysis method
Foreign crystalline compounds XRD
Cross-contamination between cathode and anode ICP-OES
Morphology and pulverization SEM
Specific surface area (m2/g) for pore formation and pulverization BET
No further work up procedures were made related to X-ray diffraction (XRD), scanning electron
microscopy (SEM) and the Brunauer-Emmett-Teller (BET) method, except conventional preparation
methods (e.g. mounting of sample and desorption of water). Analysis techniques are described in
Appendix I.
For elemental composition determination of solid samples the inductively coupled plasma optical
emission spectroscopy (ICP-OES) analysis technique was used. Samples were dissolved in aqua regia,
which was prepared from hydrochloric acid (37 wt%, Sigma Aldrich) and nitric acid (65 wt%, Merck) at
a ratio of 3:1. Complete digestion was achieved by prolonged dissolution time (minimum 18 h) and
periodical addition of hydrogen peroxide (30 wt%, Sigma Aldrich). The liquids were then heated to 70
°C for two hours and allowed to cool down. Finally, the solutions were transferred into volumetric flasks
and the volume was corrected for with pure water (>18 MΩ/cm, Milli-Q, Millipore).
The above described method was used to determine elemental concentrations of all types of samples
in this work.
Samples were diluted in 0.7 M suprapure nitric acid (65 wt%, Merck) and compared to external
calibration, prepared from standard stock solutions (1000 ppm, Ultra Scientific), using ICP-OES.
Determined metal concentrations were used as references for characterization purposes and/or
dissolution experiments.
20
Recovery of nickel was investigated by dissolution of materials in hydrochloric acid solution, with
subsequent leachate treatment carried out using solvent extraction. The experimental details are
described in the following subsections, respectively. The effect of acidity on nickel dissolution was
studied. One solvent extraction system was tested for leachate treatment by varying the O:A ratio at
two recommended extraction concentrations [93]. Determination of appropriate O:A ratio following
metal extraction was also necessary, in order to study the effect of acidity on the same extraction.
The amount of acid present in solution can affect both the dissolution of nickel and the subsequent
treatment of the leachate solution, using solvent extraction. The aspect of acid concentration was
therefore a focus for the goal of recovering Ni(s). A list of the nickel recovery experiments performed is
shown in Table 9.
Table 9. Experiments used to investigate recovery of Ni(s) from mixed electrode materials using hydrochloric acid.
Practical goal Investigation Sample
a) Dissolution rates at pH 1
Ni (s) recovery b) Characterization of undissolved Mixed electrode material
material
a) Mixed electrode material
Ni (s) recovery Nickel dissolution rates at 1-8 M HCl
b) Unspent Ni(s)
Leachate treatment Metal extraction by variation of O:A

Chloride leachate
Leachate treatment Effect of acidity on metal extraction
4.4.1 Dissolution rates of metals at pH 1

Dissolution rates of metals from mixed electrode materials (0.5 g cathode and 0.6 g anode) were
studied using hydrochloric acid solutions corresponding to pH 1. The materials were immersed into
solutions and the suspensions were stirred using plastic propellers (Metrohm). Acid was continuously
added to the solution to compensate the decrease of acidity due to dissolution of metals.
The aqueous potential was measured using a pH glass electrode (Metrohm), calibrated with buffer
solutions corresponding to pH 1, 4 and 7 (Radiometer Analytical, Metrohm). No compensation for
changes in ionic strength were made. Reported pH values should therefore only be regarded as
approximate.
The initial solution was prepared from sodium chloride salt (VWR) and the aqueous potential was
preset to pH 1 and kept constant by addition of titrant (1 M HCl). The titrant was prepared from pure
water (>18 MΩ/cm, Milli-Q, Millipore) and hydrochloric acid (37 wt%, Sigma Aldrich). Addition of
titrant was measured and controlled through a Titrando (905, Metrohm) titrator which was
connected to a computer. The aqueous potential was measured continuously and kept constant.
Samples were taken periodically in small volumes (1.5 mL), and immediately filtered (PP, 0.45 µm).
Recorded volumes and concentrations determined using ICP-OES allowed calculations of the amounts
of dissolved metal (%), with the comparison to metal concentrations in samples established previously
(section 4.2).
Characterization using XRD and SEM was carried out on undissolved mixed materials (5 g) that had been
produced by separate experiments using 5 M HCl solution, which had been terminated after 5 h. The
solid materials were washed in three steps using fresh hydrochloric acid (pH 1), followed by three steps
21
of pure water (>18 MΩ/cm, Milli-Q, Millipore) rinsing. The washed material was collected and dried in
a normal atmosphere at 55 °C until no further weigh loss could be recorded.
4.4.2 Nickel dissolution rates at 1-8 M HCl

Investigations of nickel dissolution rates from mixed electrode samples in 1-8 M HCl solutions were
carried out at room temperature (20 °C). Magnetic stir bars were used to mix the suspensions in
vials. Unspent powders of nickel metal, supplied by the battery manufacturer Nilar AB, were also
used in these experiments.
According to previous research the appropriate S/L ratio for dissolution of NiMH batteries is 1/10 [82,
83, 85, 87, 88, 90]. The amount of ligand (here chloride ion) to form metal complexes in solution is a 4
times excess for complete dissolution, assuming a 1:2 metal:ligand ratio. Samples were immersed in
acidic solution at S/L ratio of 1/20 (nickel powder) and 1/10 (mixed 0.5 g cathode and 0.5 g anode).
Acid solutions were prepared from pure water (>18 MΩ/cm, Milli-Q, Millipore) and hydrochloric acid
(37 wt%, Sigma Aldrich).
Small aliquots (0.5 mL) of samples were withdrawn periodically from the stirred solutions and
immediately filtered using a 0.45 µm filter (PP). No compensation for withdrawn volumes was made.
Nickel concentration in reference samples had been previously determined according to the procedure
in section 4.2. The amount of dissolved nickel was calculated assuming a constant S/L ratio.
4.4.3 Leachate treatment with solvent extraction

Organic phases and leachate solutions were added to vials, and the vials were sealed and shaken at
1500 vpm (vibrations per minute) using a shaking machine (IKA Vibrax VXR Basic) for 2 h. The
temperature was kept constant (25 °C) by a connected water bath. Subsequently, the vials were
centrifuged at 3500 rpm for 10 minutes (radius equal to 9 cm corresponding to 1231 g-force) with a
Labofuge 200 (Heraeus). Aqueous samples were collected before and after the extraction, and prepared
according a typical procedure (section 4.2) and analyzed using ICP-OES.
Extraction of metals from chloride leachate solutions was investigated with varied O:A ratios, using two
different concentrations of extractant Cyanex 923, with 8 vol% used to separate zinc and 70 vol% used
to extract all metals, except nickel.
Stock solutions with 8% or 70 vol% Cyanex 923 (95%, Cytec), respectively, containing the phase
modifiers 10 vol% tri-butyl phosphate (97%, Sigmal Aldrich, i.e. TBP) and 10 vol% n-decanol (99%, Sigma
Aldrich) in diluent solvent 70 (Statoil) were prepared. The organic phases were pre-equilibrated with
pure water (>18 MΩ/cm, Milli-Q, Millipore).
The chloride leachate solution was prepared by dissolution of mixed electrode material in 8 M HCl. The
acid solution was prepared with pure water (>18 MΩ/cm, Milli-Q, Millipore) and hydrochloric acid (37
wt%, Sigma Aldrich).
Acid neutralization of the 8 M HCl solution was carried out in a series of stepwise additions of mixed
electrode material. The solid-liquid dispersion was allowed to react under magnetic stirring and cooling
(ice bath) until hydrogen gas evolution could no longer be observed. The aqueous potential was
measured with a glass pH electrode (Metrohm). Additional material was added until no further change
in aqueous potential could be measured. When the potential was stable the acid neutralization process
was complete.
22
The 8 M chloride leachate solution produced was filtered (0.45 µm, PP) and sampled. The results from
ICP-OES analysis of sampled leachate are presented in Table 10.
Table 10. Concentrations (10-3 mol/L) of leachate solutions after dissolution of mixed electrode materials in 8 M HCl solutions.
The aqueous potential was measured to 310 mV.
Metal 10-3 mol/L
Al 78
Ce 60
Co 227
La 179
Mn 104
Nd 31
Ni 2902
Pr 11
Zn 57
The effect of acidity on the extraction of metals was investigated using 8 M chloride leachate solutions
at different acidities (i.e. aqueous potentials) and 70 vol% Cyanex 923. An O:A ratio of 1.4 was used and
had been previously determined as sufficient for metal extraction, but below the point of loading. Under
these conditions the metal extraction becomes dependent on the acidity of the solution.
The leachate solution presented above (Table 10) was further diluted in steps using the same stock
solution of 8 M HCl that had been used for the leachate preparation. Between each step the aqueous
solution was sampled and the aqueous potential was measured with the same electrode mentioned
above.
4.5 Recovery of anode material (including AB5 alloy)

Recovery of reusable anode material was investigated by dissolution experiments where 4 carboxylic
acids (malonic, maleic, acetic and citric acid) were used. Here, the main goals were to chemically
separate the anode material from the cathode material by dissolution of the cathode, and to remove
corrosion products from the anode.
To be precise, anode samples consist of both AB5 alloy and Ni(s), which can potentially be reused as a
mixture. Corrosion products only need to be removed from the AB5 component in order to reuse the
AB5 and Ni(s) mixture as anode material.
Chemical separation relies on different dissolution rates of the cathode and anode. There are no
theoretical indications that the AB5 alloy is immune to oxidation in acid solutions. Determining
dissolution rates of both cathode and anode material is therefore essential for the evaluation of
chemical separation. The results from experiments on separate material types can be used to evaluate
the dissolution of electrode mixtures.
Characterization of undissolved samples containing anode components is important for the

determination of selective removal of corrosion products from the AB5 alloy surface.
A list of experiments done in this work for the recovery of reusable anode is shown in Table 11, and the
details of the experiments are described in the following subsections. The same results from
experiments using 4 different carboxylic acids were used to evaluate chemical separation of
cathode/anode and removal of corrosion products. 1 M maleic acid was further chosen for separation
of cathode/anode by increasing the dissolution time and temperature, in order to increase the amount
of metal dissolution from the cathode.
23
Table 11. Experiments for the recovery of reusable anode material using carboxylic acids.
Practical goal Dissolution condition Sample Investigation 1.
cathode
4 carboxylic acids
anode
Chemical cathode
maleic acid / 10 h
separation anode Dissolution
Investigation 2.
rates
cathode
maleic acid / 50 °C anode Characterization
mixed electrodes of undissolved
Corrosion material
products 4 carboxylic acids Anode
removal
4.5.1 Dissolution experiments using carboxylic acids

Dissolution experiments were carried out on cathode and anode samples (respectively). The samples
were suspended in 1 M carboxylic acid (malonic, maleic, acetic and citric) solution at a S/L of 1/10
through magnetic stirring at room temperature (20 °C). The same experiment was also prolonged to 10
h in the case of maleic acid.
Spent cathode and anode samples, respectively, and a mixture thereof (0.5 g cathode and 0.5 anode)
were immersed in 1 M maleic acid at 50 °C for 75 minutes. The S/L ratio was 1/10 for the separate
fractions, and 2/10 for the mixed sample. The temperature was kept constant using a double-walled
glass vial connected to a heating bath, and a plastic propeller was used for stirring.
Fresh 1 M acid solutions were prepared prior (< 48 h) to the experiments by dissolving appropriate
amounts of malonic, maleic respectively citric acid salts (Sigma Aldrich) in pure water (>18 MΩ/cm, Milli-
Q, Millipore), in a volumetric flask. A volumetric flask was also used for the dilution of concentrated
acetic acid (99%, Sigma Aldrich).
The pH was measured using a pH glass electrode (Metrohm) before and after dissolution experiments
that were conducted at room temperature. Buffer solutions at pH 1, 4 and 7 (Radiometer Analytical,
Metrohm) were used for calibration. Variations of aqueous potentials due to ionic strength were not
considered, and therefore reported pH values can only be regarded as approximated values.
Small aliquots of samples (0.75 mL) were withdrawn periodically from solutions and immediately filtered
using a 0.45 µm PP-filter. Metal concentrations were determined by ICP-OES measurements, according
to procedures described in section 4.2. Dissolution of metals (%) was calculated by the assumption of
constant S/L ratio and previously determined metal concentrations of reference samples. No
compensations for withdrawn volumes were made.
In dissolution experiments for the anode samples the experiment was terminated, and undissolved
material was rinsed using respective fresh 1 M carboxylic acid in three steps. Three steps of pure water
(>18 MΩ/cm, Milli-Q, Millipore) rinsing followed. The residual material was dried in a normal
atmosphere for 24 h at 55 °C, and analyzed using XRD and SEM. EDS analysis was carried out on
undissolved mixed electrode samples from experiments at 50 °C using the same sample work up
procedure.
24
Separation of cobalt enriched surface layers from the spent cathode materials were investigated using
ascorbic acid. The feasibility of selective dissolution of cobalt was initially investigated, and most
subsequent experiments were carried out to further investigate the effects of reactions observed in the
initial experiments.
The undissolved material remaining after each experiment was characterized to evaluate the effect of
acid treatment on the anode material.
The list of experiments in Table 12 follows the order of presented results. The details of the experimental
work are described in the following subsections.
Table 12. Separation of cobalt enriched surface layers from cathode material
Ascorbic Analysis
Sample Practical goal Investigation
treatment technique
1M/4h feasibility
Dissolution
Decoupling of redox/acidic rates ICP-OES
pH 2-5
Spent properties
cathode Scale up and sample
S/L ratio
1M/2h homogeneity
non-destructive method
Morphology SEM
confirmation
Cathode
a) untreated
a) unspent Removal of cobalt surface
b) untreated Surface species XPS
b) spent enrichment
c) 1 M / 2 h
c) treated
Dissolution rate
non-destructive method ICP-OES,
Spent anode 1M/4h and residual
confirmation XRD, SEM
characterization
4.6.1 Dissolution experiments using ascorbic acid

Dissolution experiments using ascorbic acid were carried out on spent cathode and anode samples,
respectively, at an S/L ratio of 1/10. Solutions were stirred with magnetic stir bars and the room
temperature was monitored (20 °C) and kept constant. The same conditions were adopted for
experiments with varied S/L ratios or pH.
Fresh acid solutions (1-1.5 M) were prepared (< 48 h) prior to the experiment by dissolution of L-ascorbic
acid salt (Sigma Aldrich) in pure water (>18 MΩ/cm, Milli-Q, Millipore). For the investigations of the
effect of pH, 1.5 M buffer solutions of acid respective conjugate base (sodium L-ascorbate, Sigma
Aldrich) were mixed, based on approximations according to Henderson-Hasselbalch [106]:
[𝑏𝑎𝑠𝑒]
𝑝𝐻 = 𝑝𝐾𝑎 + 𝑙𝑜𝑔 [𝑎𝑐𝑖𝑑] 𝑖𝑛𝑖𝑡𝑖𝑎𝑙 Eq. 15
𝑖𝑛𝑖𝑡𝑖𝑎𝑙
Measurements of pH, sampling, calculations and further sample work up procedures for
characterization of undissolved material were similar to those previously described (sections 4.4.1 and
4.2).
25
5. RESULTS AND DISCUSSION
5.1 Characterization
Identification of crystalline compounds by XRD analysis was used as a first indication of how the
electrode materials had changed during the battery use phase. Figure 8a-b shows the cathode material,
both unspent (black) and spent (red), and anode electrode fractions, respectively. The XRD peaks of the
cathode material did not indicate the presence of any foreign crystalline phases within samples, only
phases of β-nickel hydroxide (doped with cobalt and zinc) and metallic nickel that are consistent with
JCPDS Card No: 00-059-0459 [125] and 01-070-1849 [126], respectively. No difference was noticed
between the spent and unspent cathode samples. According to the manufacturer, these samples
contain both metallic nickel and cobalt. Due to the similar scattering properties of these metals it was
not possible to differentiate between these elemental phases.
Results from XRD analyses of unspent and spent anode samples can be seen in Figure 2b. The two phases
that were identified in unspent samples correspond to metallic nickel and AB5 alloy, due to the
consistency with JCPDS Card No: 01-070-1849 [126] and 04-009-7537 [127], respectively. The spent
anode sample showed a similar diffraction pattern, however the corrosion product that corresponds to
REE(OH)3 (e.g. JCPDS Card No: 01-074-0665 [128]) was also detected. It is reasonable to assume that
the corrosion product includes a mixture of four REEs (La, Ce, Nd and Pr). These elements are chemically
alike due to the lanthanide contraction phenomena [26], and all these REEs are used in the alloy
formulation.
26
[a] Cathode

Spent Unspent
 Ni
 ß-Ni hydroxide






   
 
20 30 40 50 60 70 80
2
[b] Anode
 Spent Unspent


 Ni
 AB5 alloy
 REE(OH)3
  


 


   
   
 
    
20 30 40 50 60 70 80
2
Figure 8. XRD lines for (a) cathode and (b) anode. Both spent (red) and unspent (black) diffraction patterns are shown. X-ray
diffraction spectral lines of (a) unspent (red) and spent (black) cathode samples and (b) unspent (red) and spent (black) anode
samples. The suggested phases Ni, β-nickel hydroxide, AB5 alloy and REE(OH)3 are consistent with JCPDS Card No: 01-070-1849
[126], 00-059-0459 [125], 04-009-7537 [127] and 01-074-0665 [128], respectively.
SEM analysis was done to observe changes in particle size, shape and morphology of both electrode
materials during battery cycling. Cathode particles appeared to not have been changed by battery
cycling, as can be seen by comparing SEM images of unspent and spent samples in Figure 9a-b,
respectively. Independent of sample type (unspent and spent) there was a mixture of completely
spherical and cracked and shattered particles.
27
[a] [b]
Figure 9. SEM images of (a) unspent and (b) spent cathode material.
AB5 alloy particles from unspent and spent anode samples are shown in Figure 10. AB5 particles in spent
samples were generally smaller, and had crevices, compared to equivalent unspent particles.
Furthermore, on the AB5 alloy particles a needle shaped configuration could be detected, as can be seen
in Figure 11. Based on the literature and the XRD analysis these compounds are identified as REE(OH)3,
which formed as a result of alloy corrosion in the electrolyte [69, 71, 72].
[a] [b]
Figure 10: SEM images of (a) unspent and (b) spent anode material.
28
Figure 11. High resolution SEM image of spent anode material.
The metal composition of both spent electrode materials as determined by ICP-OES is shown in
Table 13. In the cathode nickel, cobalt, zinc and low concentrations of aluminum and manganese were
observed. Aluminum and manganese were not detected in the equivalent unspent cathode samples and
are believed to be present in the spent cathode due to cross contamination from the anode. This is not
unexpected and contaminants on the spent cathode material have been observed by others [23, 37].
For the spent anode samples, no zinc could be detected.
Table 13. Composition of spent cathode and anode electrode samples.

Spent Cathode Spent Anode
mg·g-1 mg·g-1
Al 2.1 ± 0.1 14.5 ± 0.1
Ce 44.4 ± 0.3
Co 41.9 ± 0.1 55.0 ± 0.4
La 166.2 ± 1.0
Mn 0.5 ± 0.1 37.0 ± 0.2
Nd 28.7 ± 0.2
Ni 605.2 ± 2.1 564.9 ± 3.7
Pr 10.3 ± 0.1
Zn 20.7 ± 0.1
Total 670.4 ± 2.5 921.0 ± 6.0
The detected metal content of the cathode (670.4 mg/g) is an expected result, as the main components
are metal (Ni, Co and Zn) hydroxides. If the cathode reference sample would only consist of pure nickel
hydroxide, the amount of nickel should correspond to 633 mg/g. The deviation from this value (37.4
mg/g) is probably most affected by the presence of Ni(s) powder that has been mixed into the cathode
electrode.
The degree of oxidation of the spent anode appears to be low as only 79 mg/g of the reference sample
weight was not accounted for, which indicates that most of this weight fraction is oxygen. The amount
of oxidation of the anode can be approximated to 5.5 wt% based on the following assumptions:
homogeneous oxidation of elements, their common oxidation states and the formation of only
hydroxides. This approximation is, however, dubious.
The surface area of both materials was determined using BET measurements. The results from these
measurements are shown in Table 14.
29
Table 14. Surface area of electrode samples using BET method kept isothermal at 77 K and with N2.
Electrode Battery Surface Area (m2·g-1) Increase Factor
Unspent 6.12 ± 0.10
Cathode 4.6
Spent 28.34 ± 8.28
Unspent 0.20 ± 0.01
Anode 12.3
Spent 2.42 ± 0.91
The cathode surface area increased from 6.12 to 28.34 m2 g-1 from the unspent to spent battery. The
nickel hydroxide material can undergo several phase transitions (α, β, γ) that result in increased surface
area, leading to pulverization and/or pore formation. Therefore, this is an expected result.
For the anode material the specific surface area had increased from 0.20 to 2.42 m2 g-1. This is expected,
because pulverization of the anode material is expected to have occurred with changes in lattice volume
during intercalation of hydrogen into the AB5 alloy lattice [23].
From the results of the characterization presented here some conclusions can be drawn about the
degradation of the two spent electrode materials. Degraded qualities will effect strategies to recycle the
materials.
For the cathode material it was not possible to detect any foreign crystalline compounds in the samples,
which indicates that the material has not been changed by the battery cycling. However, low amounts
of the contaminants aluminum and manganese were detected in the analysis of elemental composition.
The presence of these two elements can affect the preferred cycling between β-Ni(OH)2/β-NiOOH. BET
measurements showed that the cathode had degraded through increased surface area. A change in void
size distribution and/or pore formation can affect electrolyte diffusion, which indirectly can cause
increased internal resistance. Thus, charge/discharge characterisistics are degraded. Analysis with SEM
did however not raise suspicions of severe size or morphology changes during cycling.
In the anode material the original AB5 alloy and Ni(s) crystal structures were detected using XRD, and
also a foreign REE(OH)3 phase. Through SEM analysis this phase was observed as needle-shaped
configurations on the AB5 alloy surface. Several different corrosion scales that reside underneath these
needles have been observed by others [69, 71, 72]. All corrosion products can interfere with the
functions of the alloy surface. Specifically, the functions are to catalytically react with the electrolyte
and transport hydrogen to and from the alloy bulk [21]. Thus, it is only reasonable to assume that
corrosion products such as REE(OH)3 affect the charge/discharge rate of the anode electrode.
The elemental composition of the anode was analyzed using ICP-OES. The experimentally calculated
weight of metals in the spent anode material was close to 1000 mg·g-1, which suggested that only a part
(79 mg·g-1) is oxygen and/or hydroxide. No contamination of zinc could be observed on the spent anode
material. The amount of corrosion products was not substantial, and was estimated to be approximately
≈ 5.5 wt%.
An increase in surface area was noted for the spent anode material using the BET method. Increased
surface area will increase both preferred and parasitic reactions on the AB 5 alloy surface. Removal of
corrosion products from the spent AB5 alloy can potentially produce a material that can be used in
applications where high discharge-rates are preferred. Therefore, it is interesting to further investigate
treatment methods that recover reusable AB5 alloy from spent anode materials.
30
5.2.1 Effect of acidity on nickel recovery
Results from dissolution experiments of mixed electrode materials at pH 1 are shown in Figure 12a, and
are in agreement with those obtained by Larsson et al. [113]. From analysis of elemental composition >
90% of Al, Co, Mn, La and Zn were dissolved after 5 h. At this time less (80%) of the nickel was dissolved.
After 30 h of dissolution the amount of dissolved Al, Co, Mn, La and Zn had not changed much compared
to the change in the amount of dissolved nickel (100%). This indicates that nickel is mainly dissolving
during the period between 5-30 h. Lanthanum is only found in the AB5 alloy and zinc is only found in the
nickel hydroxide. Therefore one can expect that the dissolution rate of AB5 alloy > Ni(OH)2 (s) in
hydrochloric acid solutions at pH 1.
According to the added titrant rate (Figure 12b) the dissolution rate of the waste appears to be divided
into a fast (0-5 h) and a slow regime (5-30 h). These two regimes have very different dissolution rates
and indicate that the different components AB5 alloy, nickel hydroxide and Ni(s) are dissolving at
different rates.
The dissolution rate of the remaining 17% nickel (approximately) after 5 h acid treatment is slower
compared to those of the other elements. Nickel exists in mixed electrode materials in three main forms,
namely Ni(s), AB5 alloy and nickel hydroxide. It is expected that Ni(s) is only a small fraction of the
samples. Therefore, the 17% nickel that is dissolving slowly is attributed to Ni(s). This is a reasonable
conclusion since hydrogen desorbs slowly from nickel surfaces.
[a] Mixed electrodes [b] Mixed electrodes

32.5
100
30.0
27.5
80 25.0
Metal dissolution (%)
Volume added (mL)
Al
22.5
Co
20.0
60 La
17.5
Mn
15.0
Ni
40 Zn 12.5
10.0
7.5
20
5.0
2.5
0 0.0
0 2 4 6 24 26 28 30 0 10 20 30 40 50
Time (hours) Time (hours)
Figure 12. Dissolution of 1.1 g mixed electrode fraction in pH 1 HCl solution at 25 C. The results show (a) dissolved amount of
elements (%) and equivalent (b) 1 M volume acid addition to maintain pH.
Undissolved material after 5 h of dissolution of mixed electrode materials in hydrochloric acid at pH 1

was characterized. An interesting observation during the practical work was that the undissolved
material was magnetic and could easily be collected with the use of a magnet. The nickel content of
these collected residuals was high (96.6 + +
− 1.0 wt%) and they also contained some cobalt (0.3 − 0.1 wt%).
No other elements could be quantified ≥ 0.1 wt%. The unaccounted amount (≈ 3 wt%) is believed to be
related to oxygen present. Analysis results from XRD and SEM can be seen in Figure 13 and Figure 14,
respectively.
31
Undissolved mixed electrode residue

sample repetition
1
2
 Ni 3
20 30 40 50 60 70 80
2
Figure 13. Results from XRD analysis of residual material collected from three replicates (red, black and blue) obtained from
dissolution of mixed electrode material, using pH 1 hydrochloric acid solutions at 25 C, after 5 h. The suggested phase Ni is
consistent with JCPDS Card No: 01-070-1849 [126].
Figure 14. SEM images using Quanta 200 FEG ESEM of unused nickel powder (left) and spent nickel powder (right) collected after
dissolution experiments at pH 1.
Results from dissolution of mixed electrode materials in 1-8 M hydrochloric acid solutions are shown in
Figure 15a-b. The dissolution rates of the AB5 alloy and Ni(OH)2 (s) components are illustrated with the
amount of dissolved lanthanum and zinc, respectively. Dissolution rates of these two components are
complete after 0.5 h in 2-8 M hydrochloric acid solutions. The corresponding dissolution rate of nickel
can be seen in Figure 15b. It is an expected result that the dissolution rate of nickel is different compared
to that of lanthanum and zinc, since some of the nickel originates from the Ni(s) component.
The dissolution rate of nickel from pure Ni(s) powder samples in hydrochloric acid solutions (1-8 M) can
be seen in Figure 16. Interestingly, the dissolution rate of nickel in 1-4 M hydrochloric acid is similar and
in 8 M solution almost complete dissolution of nickel is achieved after 5 h immersion time.
32
[a] Mixed electrodes [b] Mixed electrodes
100 100
80 1M La 80

Zn
Nickel
60 2M La 60 1M
Zn 2M
4M
40 40
4M La 8M
Zn
20 20
8M La
Zn
0 0
0 1 2 3 4 5 0 1 2 3 4 5
Figure 15. Dissolution of (a) lanthanum, zinc and (b) nickel from 0.5 mixed electrode material initially in 1-8 M hydrochloric acid
solutions.
Pure Ni(s) powder

100
Nickel
80 1M
2M
4M
60
8M
40
20
0
0 1 2 3 4 5
Time (hours)
Figure 16. Dissolution of nickel from 0.25 g pure Ni(s) powder samples initially in 1-8 M hydrochloric acid solutions.
Recovery of Ni(s) can be made by dissolving mixed electrode materials in hydrochloric acid. In a practical
recycling process the mixed electrode materials are immersed in e.g. 8 M hydrochloric acid in order to
dissolve the materials. The dissolution rate of components follows the decreasing order AB5 alloy ≈
Ni(OH)2 (s) >> Ni(s) in 2-8 M hydrochloric acid concentration and AB5 alloy > Ni(OH)2 (s) >> Ni(s) in 1-0.1
(i.e. pH 1) M acid concentration. That is to say, the acid is rapidly consumed by the AB5 alloy and Ni(OH)2
(s) components. Therefore one should expect to recover the majority of Ni(s) powder component even
when mixed electrode materials are immersed in 8 M hydrochloric solutions.
33
5.2.2 Extraction of metals using Cyanex 923 and the effect of acidity
When the logarithmic distribution ratio of a metal exceeds zero the metal is more concentrated in the
organic phase compared to the aqueous phase. The separation of one metal from others is increased
when the differences in distribution ratios between the metals are increased.
The distribution ratio of zinc from the leachate solution increases with increasing O:A ratio, as can be
seen in Figure 17. The distribution ratio of zinc is at least 1.75 higher order of magnitude compared to
distribution ratios of the other metals (Al, REE, Mn and Co) at O:A = 3. Thus, zinc can be separated from
the other metals. The extraction and separation of zinc from similar leachate solutions, by keeping a low
concentration (8 vol%) of Cyanex 923, is in agreement with the results of others [93].
Variation of general metal extraction with respect to O:A ratio can be seen in Figure 18. All metals except
nickel were extracted, because the concentration of Cyanex 923 was high (70 vol%). Above a 3.5 O:A
ratio the extraction of metals does not increase. As anticipated, Cyanex 923 can be used to extract and
separate metals from nickel [93].
8 vol% Cyanex 923

2.0 Al
Ce
1.5 Co
La
1.0 Mn
Nd
0.5
Pr
Log D
Zn
0.0
-0.5
-1.0
-1.5
-2.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0
O:A
Figure 17. Extraction of zinc and other metals from 8 M chloride leachate solution using 8 vol% Cyanex 923 in Solvent 70. Nickel
was not extracted.
70 vol% Cyanex 923

2
Al
Ce
Co
La
Mn
1
Nd
Pr
Log D
-1
1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
O:A
Figure 18. Extraction of metals from 8 M chloride solutions using Cyanex 923 (70 vol% in Solvent 70). Zinc was quantitatively
extracted and nickel was not extracted.
34
The results from experiments at different aqueous potentials can be seen in Figure 19. For the sake of
reference, 440 mV corresponds to approximately 0.12 M HCl concentration. At a measured aqueous
potential up to 440 mV the extraction of all metals does not vary with the amount of HCl concentration.
The Log (DLa = -0.25) corresponds to approximately 40% extraction of lanthanum. Above 440 mV the
extraction of aluminum and REEs decreases. It was described by Larsson et al. [93] that the extraction
of REEs and aluminum benefits from a high chloride concentration but not from increased acidity. The
amount of available extracting agent (Cyanex 923) that can extract REEs and aluminum are instead
consumed by the extraction of acid. Therefore, the results obtained from these experiments are
expected.
70 vol% Cyanex 923 vs acidity

2.5
Al
2.0 Ce
Co
La
1.5
Mn
Nd
1.0
Pr
Log D
Zn
0.5
0.0
-0.5
-1.0
-1.5
300 320 340 360 380 400 420 440 460 480 500
Aqueous potential (mV)
Figure 19. Extraction of metals from 8 M chloride solutions, at varied aqueous potentials. An 8 M chloride solution has been
diluted with 8 M hydrochloric acid, which resulted in an increase of aqueous potential registered with a glass electrode. The
organic phase consists of Cyanex 923 (70 vol% in Solvent 70). The experiments were carried out at a constant O:A ratio of 1.4.
Experimental results from varied O:A ratio experiments confirm the Cyanex 923-chloride system as
appropriate for the separation of zinc, aluminum, cobalt, manganese and REEs from nickel.
The experimental results presented here demonstrate that controlling the acidity is important for
maintaining extraction of REEs and aluminum from 8 M chloride solutions using Cyanex 923. A
concentration of 0.12 M HCl was identified as a threshold value for acceptable REE and aluminum
extraction at 70 vol% Cyanex 923. This threshold implies that 15 mL 8 M HCl is the tolerated level of acid
overshooting when producing 1 L of leachate solution. This accuracy (15 mL) can be achieved on an
industrial recycling scale. However, further work is needed to determine the target level.
35
5.3 Recovery of reusable anode using carboxylic acids
5.3.1 Chemical separation of anode and cathode
A summary of the amounts of dissolved metals from cathode respectively anode materials after 4 h
treatment in carboxylic acids (acetic, citric, maleic and malonic acid) are shown in Figure 20. From these
results one can expect no chemical separation of the electrode materials to be achieved without losses
of anode. Possibly, the investigated dissolution period (4 h) was not sufficient for complete dissolution
of the cathode material.
pH 2-3 pH 2 pH 1 pH 1-2
100
80
60
40
20
0
Al Co La Mn Ni Co Ni Zn Al Co La Mn Ni Co Ni Zn Al Co La Mn Ni Co Ni Zn Al Co La Mn Ni Co Ni Zn
Anode Cathode Anode Cathode Anode Cathode Anode Cathode
Acetic Citric Maleic Malonic
Figure 20. Summary of the amount dissolved of the main constituent metals (%) in the electrode materials in the respective 1 M
carboxylic acids, after 4 h dissolution at room temperature (20 °C). The pH values before and after the experiments were similar
and approximated with the indicated values.
In trial experiments at 50 °C only maleic acid appeared useful (between the 4 carboxylic acids) with
respect to achieving separation of cathode and anode material, and maleic acid was selected for further
experimental tests.
Dissolution rates of metals from cathode and anode material in 1 M maleic acid solutions are shown in
Figure 21a-b, respectively. The amount of dissolved metals from cathode as well as from anode
increases with time. After 10 h the dissolution of the cathode material was still not complete. However,
it is indicated in Figure 21a by the dissolution curve that the amount of metals released from the cathode
appear to converge. Therefore, separation of cathode and anode is possibly coupled to a limited
solubility of zinc, cobalt and nickel.
Dissolution of all metals from anode samples was high (≥ 30%) after 10 h, as can be seen in Figure 21b.
This indicates that the AB5 alloy material was dissolving.
36
[a] Cathode [b] Anode
100 100
Co Al
Ni Co
80 Zn 80 La

Mn
Ni
60 60
40 40
20 20
0 0
0 2 4 6 8 10 0 2 4 6 8 10
Figure 21. Dissolution of 0.5 g (a) anode and (b) cathode samples in 50 mL 1 M maleic acid solutions at 20 °C.
The effect of elevated temperature (50 °C) on the dissolution rates of metals from cathode and anode
in 1 M maleic acid is shown in Figure 22a-b, respectively. With respect to room temperature (20 °C), the
amount of dissolved metals from both cathode and anode were increased.
Zinc had been dissolved completely (100%) after 60 min in 1 M maleic acid at 50 °C, which is an
important observation to make from the results shown in Figure 22a. The amount of dissolved nickel
and cobalt are approximately the same (70%). The current dissolution experiment can potentially still
constitute a separation of the cathode and anode material. Nickel exist as powder, and cobalt exists as
both powder and cobalt-enriched surface layers. The amount of these three compounds in the cathode
samples have not been determined. However, these should only exist at low amounts. Thus, the
remaining undissolved 30% of nickel and cobalt can possibly be attributed to these three components.
Based on these assumptions it is indicated that the nickel hydroxide has reacted completely with maleic
acid.
It can be estimated that 50% of the AB5 alloy had been dissolved from the anode sample immersed in 1
M maleic acid at 50 °C, as shown in Figure 22b. A lower fraction of the nickel was dissolved (40%). This
is expected, because metallic nickel powder also exists in the anode samples.
[a] Cathode [b] Anode

100 100
Co Al
Ni Co
80 Zn 80 La
Mn
Ni
60 60
40 40
20 20
0 0
0 10 20 30 40 50 60 70 80 0 10 20 30 40 50 60 70 80
Time (minutes) Time (minutes)
Figure 22. Dissolution of 0.5 g (a) anode and (b) cathode samples in 50 mL 1 M maleic acid solutions at 50 °C.
Samples treated with 1 M maleic acid at 50 °C for 75 min were characterized with XRD, and the results
are shown in Figure 23. In agreement with the conclusion drawn from the dissolution investigation
(Figure 22a-b) diffraction peaks corresponding to Ni(OH)2 (s) could not be detected. Instead, only
crystalline phases corresponding to anode constituent materials AB5 alloy and Ni(s) were detected.
37

 Spent and treated samples:
cathode
mixed cathode and anode
 Ni
 AB5



      

 
 
20 30 40 50 60 70 80
2
Figure 23. Results from XRD analysis of collected solid residual material from dissolution experiments with 1 M maleic acid
solution at 50 °C for 75 minutes. Note that the diffraction peaks corresponding to the anode components AB5 alloy and Ni(s)
originate from a mixed cathode and anode sample. The phases Ni and AB5 alloy are consistent with JCPDS Card No: 01-070-1849
[126] and 04-009-7537 [127], respectively.
A SEM image and EDS map of mixed cathode and anode sample that had been treated with 1 M maleic
acid for 75 min at 50 °C is shown in Figure 24. The coupled SEM/EDS analysis illustrates that patch-wise
corrosion layers (purple) rich in nickel (red) and carbon (blue) covered the AB 5 alloy particles. These
layers possibly consist of nickel complexes that were not dissolved during dissolution experiments.
There are no signs of REE(OH)3 needle formation on the AB5 alloy particles. Signs of pitting corrosion
were generally observed on the AB5 alloy particles.
Figure 24. SEM images taken with Ultra 55 FEG SEM showing an anode particle separated from mixed electrode material, using
1 M maleic acid at 50 °C for 75 min. Images (left and right) are of the same surface area from one particle and illustrate patch
wise surface layers. EDS analysis uses the following colors: nickel (red), oxygen (green), carbon (blue) and nickel-carbon (purple).
An industrial and practical method, using carboxylic acids, does not appear to be feasible for the
chemical separation of cathode and anode. These acids would only be practically useful in the case
where the dissolution behaviors (rates and amounts) of cathode and anode are very different in the
same solution. Such differences in properties were not observed for any of the investigated carboxylic
acids (malonic, maleic, acetic and citric acid).
It was demonstrated that increasing the temperature from 20 °C to 50 °C for the 1 M maleic acid
solution could achieve a chemical separation of cathode and anode material. However, the anode was
38
significantly dissolved (50%). Dissolution of AB5 alloy is coupled to the electrochemical oxidation where
hydrogen ions are reduced and hydrogen gas evolves. This is a general drawback for the chemical
separation of NiMH electrode materials using acidic solutions.
5.3.2 Separation of corrosion products from AB5 hydrogen storage alloy

Dissolution rates of elements from anode samples using 1 M carboxylic acids (malonic, maleic, acetic
and citric acid) are shown in Figure 25 and Figure 26, respectively. The dissolution of the main anode
constituent elements nickel and lanthanum decreases in the order of malonic > maleic > acetic ≈ citric
acid. The practical pH of the solutions did not vary before and after the experiment, and has been
presented previously in Figure 20.
As previously pointed out it can be roughly approximated that 5.5 wt% of the sample was corrosion
products, i.e. lanthanide hydroxides. Due to the high (> 50%) amounts of dissolved metals it is therefore
indicated that the AB5 alloy was dissolving as well in the malonic acid solutions.
The amounts of dissolved metals from anode material were different with respect to the different
solutions used (maleic, acetic and citric acid), as shown in Figure 25a-b and Figure 26a-b. The amount
of dissolved metals into maleic acid were low (< 20%). Interestingly, the amount of dissolved metals into
acetic and citric acid was even lower (< 10%), and also appeared to converge. Previous investigations of
the corrosion products on the spent AB5 alloy surface suggest that many different oxidized compounds
(Ni, Co, REE, Mn) can exist. Therefore, the observed different dissolution rates indicate that some of the
corrosion products may have been removed.
[a] Malonic acid [b] Maleic acid

100 20
Al Al
Co Co
80 La La
Mn 15 Mn
Metal dissolved (%)
Metal dissolved (%)
Ni Ni
60
10
40
5
20
0 0
0 50 100 150 200 250 0 50 100 150 200 250
Figure 25. Dissolution of 0.5 g anode material in 50 mL 1 M (a) malonic and (b) maleic solutions.
[a] Acetic acid [b] Citric acid

20 20
Co Co
La La
Mn Mn
15 15
Metal dissolved (%)
Metal dissolved (%)
Ni Ni
10 10
5 5
0 0
0 50 100 150 200 250 0 50 100 150 200 250
Figure 26. Dissolution of 0.5 g anode material in 50 mL initially 1 M (a) acetic and (b) citric acid solutions.
39
The results from the characterization of the collected anode samples (subjected to 4 h treatment in
malonic, maleic, acetic and citric acid) using XRD can be seen in Figure 27a. For all samples the same
crystalline phases corresponding to the AB5 alloy and Ni(s) were detected. For the sake of comparison,
the highest relative intensities of REE(OH)3 phases, which occur between 27-29, 39-41 and 48-50 2θ
angles are shown in Figure 27b. As far as can be seen from XRD analysis the REE(OH)3 has been removed
by the treatment of malonic, maleic, acetic and citric acids.
[a] Anode samples treated with acid:

Spent (untreated)
Maleic 

Malonic   Ni
Citric  AB5
Acetic  REE(OH)3
 



 
 
   
   
 

  

15 20 25 30 35 40 45 50 55 60 65 70 75 80 85
2
[b] Anode samples treated with acid:

 REE(OH)3 Spent (untreated)
Maleic
 Malonic

Citric
Acetic
 
27 28 29 39 40 48 49 50
2
Figure 27. XRD diffractograms of spent and untreated anode and anode samples subjected to 4 h treatment in different 1 M
carboxylic acids (a). The resolution of relative intensities at some 2Θ intervals has been increased (b). The phases Ni, AB5 alloy
and REE(OH)3 are consistent with JCPDS Card No: 01-070-1849 [126], 04-009-7537 [127] and 01-074-0665 [128], respectively.
SEM analysis was conducted on anode samples that had been treated with 1 M carboxylic acids for 4 h
and the results are shown in Figure 28a-b (malonic acid) and Figure 29a-b (acetic respectively citric acid).
The effect of treatment with maleic acid that can be expected has been previously presented in Figure
24. The images that are shown here represent the most important observations that were made during
analysis.
40
In all analyzed samples the characteristic irregular anode particle shape that resembles the shapes of
unspent AB5 alloy particles was detected. This indicates that neither treatment completely destroyed
the particle shape.
In samples subjected to the aggressive leaching solution (malonic acid) the surfaces of particles were
visually alike. Particles subjected to malonic acid appeared to have been subjected to uneven corrosion,
e.g. pitting corrosion (Figure 29a). In addition, the surfaces of the examined particles had patch-wise
surface layers (Figure 29b). In comparison to the EDS analysis of samples treated with maleic acid at
50 °C (Figure 24) these patch-wise surface layers are visually similar. Therefore it is reasonable to
assume that the patch-wise surface layers seen here are also nickel-carbon enrichments.
[a] [b]
Figure 28. SEM images taken with Ultra 55 FEG SEM. Anode particles treated with 1 M malonic acid for 4 h. In the left hand Figure
(a) patch-wise corrosion products are observed. In the right hand Figure (b) pit formation (i.e. pitting corrosion) is seen.
For the least aggressive treatment media (acetic and citric acid) the visual appearances of particles were
similar regardless of treatment solution. Most of the AB5 alloy particles in these samples were covered
with irregular disc-like crystals (see Figure 29a-b). These surfaces do not resemble the surface of
unspent AB5 alloy. It is therefore believed that the observed surface layers are acetate or citrate metal
complexes, respectively, or remaining undissolved corrosion products from battery cycling (i.e.
nickel/cobalt oxides).
[a] [b]
Figure 29. SEM images taken with Ultra 55 FEG SEM of anode particle treated with (a) acetic and (b) citric acid for 4 h. The surface
morphologies of observed particles were comparably similar regardless of the acid (acetic and citric) used to treat the anode
samples. The image to the right (b) is at high resolution.
It was not possible to recover an AB5 alloy material completely free from inorganic and/or organic
corrosion product from spent anode material using the carboxylic acids chosen for the current work.
41
Carboxylic acids contain oxygen in the functional group (COOH) that can interact with the surfaces of
the solid samples. These interactions can result in kinetically slow surface processes. In fact, the
dissolution rates of the anode material (in carboxylic acids) presented in the current work are generally
low compared to previous investigations (in mineral acids) [113]. Therefore, the attempt to use
carboxylic acids to slow down the dissolution of AB5 was to some extent successful.
The practical usefulness of the investigated carboxylic acids is, however, limited. An interesting
conclusion that can be drawn is that acetic and citric acid can be used to remove the REE(OH)3
compounds from the surface without significantly dissolving the AB5 bulk. This was not observed for the
two other acids (malonic and maleic acid). From electrochemical considerations it is predicted that the
AB5 alloy will eventually also react with acetic and citric acid, due to the hydrogen evolution potential.
Therefore, it is probable that the relatively selective removal of REE(OH)3 is associated with slow
dissolution rates of the remaining corrosion products.

5.4.1 Cathode material
5.4.1.1 Leaching of cathode
The results from preliminary dissolution experiments in 1 M ascorbic acid are shown in Figure 30.
Dissolution of cobalt was approximately 63% after 60 minutes. At this time the dissolution of nickel
(1.4%) and zinc (3.1%) were both low. Relatively selective dissolution of cobalt using ascorbic acid was
expected, due to the reducing properties of ascorbic acid and the elevated oxidation state of cobalt at
the cathode surface.
It is difficult to determine the cause of selective dissolution of cobalt in these experiments. Interestingly
only a fraction of the cobalt was dissolved. The total concentration of cobalt in cathode material is low
(4.2 wt%) and it is suspected that cobalt exists in the following three different components: elemental
cobalt powder, cobalt hydroxide and surface enriched cobalt in oxidation states higher than 2. It is
therefore difficult to estimate the amount of cobalt that exists on the surface of the nickel hydroxide
material. Since the cathode material is complex, further experiments are needed to elaborate on the
observed selective dissolution phenomena.
Ascorbic acid
100
Cobalt
Nickel
80 Zinc
Metal dissolved (%)
60
40
20
0
0 50 100 150 200 250
Time (minutes)
Figure 30. Dissolution of the main elements cobalt, nickel and zinc from cathode samples using 1 M ascorbic acid solutions, at 20
°C and S/L ratio of 1/10.
The effect of pH on the dissolution of metals from cathode samples was investigated, and the results
have been summarized in Figure 31 and Figure 32. Aluminum and manganese that had been observed
in spent cathode samples were also considered in these experiments. The results show that the solution
pH has only a slight effect on the dissolution rate and amount (60-70%) of dissolved cobalt. The effect
42
of pH on the dissolution of other elements (Al, Ni and Zn) was similar, but the amount of dissolved
metals was generally low (> 2-4.5%). Interestingly, the dissolution behavior of manganese is similar to
that of cobalt with one exception; manganese dissolves completely (100%).
The hypothesis that a redox reaction responsible for selective cobalt dissolution was a general
statement in section 4.5. The hypothesis was confirmed by decreasing the acidity of leaching solutions
in which ascorbic acid still had a negative reduction potential.
A variation in pH did not have a significant effect on the dissolution of nickel hydroxide particles, as can
be seen in Figure 31. Nickel and zinc are constituents of the spent nickel hydroxide bulk particles, and
the dissolution of these metals was low (> 2-4.5%). Other researchers suggest that aluminum is
intercalated into the nickel hydroxide bulk [37, 69, 74, 75]. In the current results (Figure 31) it is also
suggested that aluminum is intercalated in the nickel hydroxide bulk, since aluminum dissolves with
similar rates to the bulk-constituent metals nickel and zinc.
Both cobalt and manganese dissolved fast, and manganese completely, in ascorbic acid. Thus, it is
reasonable to assume that both these metals exist at the surface of the cathode material. The reduction
potential of the pairs Mn3+/Mn2+ (1.54 V) and Co3+/Co2+ (1.92 V) are both positive [107], and the
reduction potential of ascorbic acid is negative in acidic solutions. Dissolution of these metals is
therefore thermodynamically reasonable.
[a] Aluminum [b] Nickel

5 5
pH 2 pH 2
pH 3.5 pH 3.5
4 pH 5 4 pH 5
Dissolved aluminum (%)
Dissolved nickel (%)
3 3
2 2
1 1
0 0
0 20 40 60 80 100 120 0 20 40 60 80 100 120
[c] Zinc
5
pH 2
pH 3.5
4 pH 5
Dissolved zinc (%)
0
0 20 40 60 80 100 120
Time (minutes)
Figure 31. Dissolution of (a) aluminum (b) nickel and (c) zinc in 50 mL 1.5 M ascorbic acid/sodium hydrogen ascorbate solution at
20 °C, from 0.5 g cathode material. The acidities (pH) of the initial and final solutions were similar and have been approximated
to the indicated pH values.
43
[a] Cobalt [b] Manganese
100 100
Dissolved manganese (%)

Dissolved cobalt (%)
80 80
pH 2
pH 3.5
60 60 pH 5
40 40
pH 2
pH 3.5
20 pH 5 20
0 0
0 20 40 60 80 100 120 0 20 40 60 80 100 120
Figure 32. Dissolution of (a) cobalt and (b) manganese in 50 mL 1.5 M ascorbic acid/sodium hydrogen ascorbate solution at 20
°C, from 0.5 g cathode material. The acidities (pH) of the initial and final solutions were similar, and have been approximated to
the indicated pH values.
Results from investigation of S/L ratio on the dissolution of cobalt and manganese are shown in Figure
33. Both the amount of manganese and the targeted cobalt are homogeneously distributed in the
cathode material, as indicated by the similar amount of dissolved material. Existence of manganese in
spent cathode material is a result of AB5 alloy corrosion. It is therefore only logical that also manganese
resides at the surface of the nickel hydroxide particles.
Cathode
100
80
Metal dissolved (%)
60
Co
Mn
40 Ni
Zn
20
0
0 100 200 300 400 500
solid to liquid (kg/L)
Figure 33. Dissolution cobalt, manganese, nickel and zinc in initially 1 M ascorbic acid solutions for 2 h at 20 °C, while varying the
solid to liquid ratio (kg cathode/Liter solution).
With the results presented above it is indicated that ascorbic acid is useful for producing a concentrated
cobalt solution from NiMH electrode waste. However, there are yet some aspects that need to be
further investigated. In solutions with the highest amounts of dissolved metals (249 mmol/L Co, 10
mmol/L Mn, 56 mmol/L Ni and 3 mmol/L zinc) precipitation occurred during storage of the solution.
Elemental analysis of the residual and filtered solutions showed reduced concentrations of all metals:
cobalt (30.3%), manganese (7.6%), nickel (19.3%) and zinc (27%). From a recycling point of view this is
an interesting result. It suggests that cobalt-enriched surface layers can be extracted relatively
separately and then conveniently recovered as a solid compound.
44
5.4.1.2 Characterization of undissolved cathode materials
Cathode material was analyzed with SEM after ascorbic acid treatment (1 M for 2 h) and images of these
samples are shown in Figure 34. These images show that spherical particles are still observed after
treatment with ascorbic acid. In agreement with previous characterization of unspent and spent
cathodes there are also observations of particles that have been shattered and/or cracked. Therefore
the treatment using ascorbic acid cannot be regarded as a destructive method with respect to the shape
of the particles.
Figure 34. SEM images of cathode particles from analysis using Quanta 200 FEG ESEM. Both images show particles that have been
treated with 1 M ascorbic acid for 2 h. The image to the right is at higher resolution.
A survey XPS spectrum was produced by analysis of several different cathode samples. In Figure 35 the
results from analysis of unspent and spent cathode samples, and spent cathode sample, treated with 1
M ascorbic acid (2 h) can be seen. For the current result only the indicated peak Co 2p3/2 is of interest,
which occurs at approximately 779.6 eV. The oxidation state of cobalt that is associated with this peak
is discussed in more detail in the following paragraph, in which the results of narrow scans are
presented. For now it is pointed out that this peak (779.6 eV) corresponds to certain cobalt
compound(s). The peak position shows that such cobalt compound(s) only exist on samples that have
not been treated with ascorbic acid.
Other peaks seen in Figure 35 are expected to be associated with other elements in the sample matrix
(i.e. Ni) or impurities obtained from sample work up (Si, C, Na etc.).
45
Cathode Samples:
Spent
779.6 eV Unspent
Spent and treated
1200 1000 800 600 400 200 0

Binding energy (eV)
Figure 35. XPS survey spectrum of the three different sample types spent, unspent, and spent and treated cathode samples.
Treated sample refers to treatment using 1 M ascorbic acid for 2 h.
At the immediate surface of both unspent and spent cathode particles a main peak at approximately
779.6 eV was detected, as can be seen in the narrow scan shown in Figure 36a. Binding energies at this
energy correspond to cobalt 2p2/3. According to the literature, at this binding energy (779.5 eV) several
cobalt compounds may give rise to the 2p2/3 spectral line, namely CoOOH, Co2O3 and Co3O4 [129].
Nonetheless, for these three compounds cobalt is at an elevated oxidation state, namely +2.7 (Co3O4)
or +3 (CoOOH and Co2O3). As can be seen in these narrow scans, the cobalt compounds with oxidation
states higher than 2 have been removed by treating the spent cathode samples in 1 M ascorbic acid
solution for 2 h.
Narrow acquisition scans around 860 eV for spent and unspent cathode, and treated spent cathode
samples are shown in Figure 36b. These results are presented in order to point out the presence of
nickel at the surface, which is the most abundant metal in the cathode samples. Different nickel
compounds (Ni(OH)2, NiOOH and NiO) are associated with the peak broadening around the main peak
(Ni 2p3/2, at 855.5 eV) [130]. It can be concluded that the surface speciation of nickel also varies, which
can be expected due to complex surfaces. Interestingly, the nickel surface speciation does not appear
to have changed during the ascorbic acid treatment.
46
[a] Cathode samples: [b] Cathode samples:
Spent Spent
779.5 eV Unspent 855.5 eV
Unspent
794.2 eV Treated
Treated
795 790 785 780 775 865 860 855

Binding energy (eV) Binding energy (eV)
Figure 36. Narrow scans of different cathode samples, namely spent and unspent cathode samples, and spent samples treated
with ascorbic acid (i.e. “treated”). In (a) spectral lines associated with Co 2p2/3 (779.5 eV) and Co 2p1/2 (794.2), on cathode
samples. It can be seen that the sample treated with 1 M ascorbic acid (2 h) does not have any Co 2p peaks in this range (eV). In
(b) spectral lines at around 855.5 eV correspond to Ni 2p3/2.
Detection of the contaminants aluminum and manganese was not possible with XPS. Aluminum is
difficult to detect due to its low concentration (2 mg/g) in the material and the only peaks (2s and 2p)
that could be expected in the measured range overlap with peaks corresponding to nickel 3s and 3p.
Although the manganese peak of 2p3/2 also overlaps with the nickel auger line the Mn 2p1/2 should be
detectable if the concentration is sufficiently high. However, the concentration is low (0.5 mg/g) in these
samples. Therefore it is concluded that no significant manganese enrichment on the cathode surface
could be detected using this method. Due to limited practical means of detecting manganese at such
low concentrations its existence can only be speculated.
To verify that cobalt is enriched on the surfaces of spent cathode samples it was of interest to also
analyze the depth profile. The result of this analysis is shown in Figure 37 and it can be seen that the
concentration of nickel and cobalt varies with the depth. At the immediate top surface the
concentration of cobalt is higher (25 at%) compared to deeper into the bulk, in which cobalt converges
to < 10 at%. This confirms that cobalt is more located at the surface region.
With the presented XPS results it can be concluded that the cobalt-enriched surface on spent cathode
materials can be removed by ascorbic acid. This also confirms the results obtained from dissolution
experiments. Because the cobalt-enriched surface can be selectively dissolved, the ascorbic acid
treatment may also be a practical means of evaluating the thickness of these layers.
47
Figure 37. Depth profile (at%) of spent cathode sample using Ni 2p3/2 and Co 2p1/2 binding energies (indicated by legends Ni2p3
and Co2p1, respectively). The etching rate calibrated on Ta2O5 with these conditions is 4.76 nm/min.
5.4.2 Anode material

5.4.2.1 Leaching of anode
Dissolution of anode samples in 1 M ascorbic acid solutions are presented in Figure 38. The total amount
of manganese in anode samples was 3.7 wt% and is mainly found in the AB5 alloy bulk. The observation
that only 4% of the total manganese content had dissolved suggests that only a little manganese (4% of
3.7 wt%) resides on the spent anode particle surfaces. This low amount makes it difficult to detect
manganese compounds on the surface of the spent anode material. However, according to the
observations made by Maurel et al. [72] manganese is solubilized in the electrolyte and then re-
precipitates on the alloy particle surfaces as MnO(OH) (s) and Mn(OH)2 (s). As previously pointed out,
the reduction potential of the Mn3+/Mn2+ pair is positive (1.54 V vs SHE) [107], and the reduction
potential of ascorbic acid is negative in acidic solutions. An electrochemical cell formed from manganese
(III) and ascorbic acid is therefore thermodynamically possible.
The results indicate that the physical and chemical properties of the hydrogen storage alloy particles
have not been affected by the ascorbic acid treatment.
Ascorbic acid
5
Co
La
4 Mn
Ni
Metal dissolved (%)
0
0 50 100 150 200 250
Time (minutes)
Figure 38. Dissolution of cobalt, lanthanum, manganese and nickel from anode materials using 1 M ascorbic acid solutions, at 20
°C and S/L ratio of 1/10.
48
5.4.2.2 Characterization of undissolved anode material
XRD analysis was performed on the collected undissolved anode samples from 1 M ascorbic acid
treatment for 4 h (see Figure 39). As far as can be seen from the XRD analysis, only the REE(OH)3, AB5
alloy and Ni(s) phases can be detected. This result is in agreement with the insignificant dissolution that
has been quantified through ICP-OES analysis. More importantly this result also indicates that no
additional (crystalline) oxidation products have been formed during the treatment.
Undissolved anode residue

 Ni

  AB5
 REE(OH)3
 
 


 
    
   

    

15 20 25 30 35 40 45 50 55 60 65 70 75 80 85
2
Figure 39. The figure show the relative intensities of the disposed anode material and after 4 h dissolution treatment of ascorbic
acid. The suggested phases Ni, AB5 alloy and REE(OH)3 are consistent with JCPDS Card No: 01-070-1849 [126], 04-009-7537 [127]
and 01-074-0665 [128], respectively.
SEM images of spent AB5 alloy particles treated with ascorbic acid are presented in Figure 40. In
agreement with XRD analysis it was possible to observe the REE(OH)3 needles covering particle surfaces.
Overall, the electron microscopy showed that the AB5 particles remained in their original form. The
comparison with previous SEM images (Figure 10) of unspent and spent AB5 particles indicate that the
ascorbic acid treated particles have not been completely destroyed. These results, coupled to results
obtained through XRD and ICP-OES analysis, do not raise any suspicions that the ascorbic acid treatment
destroys the AB5 alloy material.
Figure 40. SEM images of anode particles treated with 1 M ascorbic acid for 4 h, using Ultra 55 FEG SEM. To the left, three
different particle morphologies are observed. The particles in the background show two different morphologies, compared to
the particle in focus. The right images show REE(OH)3 needles at high resolution.
49
6. CONCLUSIONS
Recycling of NiMH batteries is important in order to recover economically and technologically important
metals (Ni, Co, REEs), which can be used to sustain the needs of various markets, e.g. steel industry or
new NiMH batteries. The intention of this work was to contribute to the development of sustainable
recycling methods by investigating selective recovery of the components: (1) metallic nickel, (2) reusable
AB5 alloy and (3) cobalt.
6.1 Characterization of spent NiMH batteries

The aim of characterizing NiMH materials was to evaluate the feasibility of reusing some of its
components. It was clearly shown from the characterization of spent bipolar NiMH batteries using ICP-
OES, XRD, SEM and BET that the components did not completely degrade through battery cycling.
Overall, the qualitative observations of degradation that were made are in agreement with previous
investigations.
Ultimately, the feasibility of reusing some of the spent NiMH components cannot be addressed without
practically testing new batteries made from recovered and regenerated components. Reusing the AB5
alloy material from spent NiMH batteries appears most interesting. The increased surface area can be
beneficial in further uses, and avoiding reproduction of this alloy from metal ingots reduces high
consumption of energy. However, corrosion products such as REE(OH)3 must be removed, in order to
restore charge/discharge properties.
6.2 Recovery of metallic nickel

The intention of using hydrochloric acid to treat NiMH waste had the separate goals: (1) recovery of
metallic nickel and (2) investigating compatibility of produced leachates with Cyanex 923 solvent
extraction separation of metals.
Recovery of Ni(s) powder relies on strong chemisorption of hydrogen onto nickel surfaces, thus slowing
down oxidation of nickel in acid solutions. At pH 1 the dissolution rate of Ni(s) powder is slower than
the corresponding dissolution of the AB5 alloy and nickel hydroxide components. Ni(s) powder can
therefore be recovered by using a well-chosen leaching time. This is economically favorable for two
reasons. Firstly, less acid is consumed while a relatively pure (96.6 wt%) Ni(s) powder fraction can be
recovered. Secondly, extensive time to dissolve the Ni(s) powder can be avoided.
Compatibility of chloride leachates, with different levels of acidity, with Cyanex 923 was investigated.
The results of this work re-confirmed that Cyanex 923 is useful for the separation of metals (Zn, La, Ce,
Pr, Nd, Al, Mn, Co) from nickel, as has been stated previously by Larsson et al. [93]. It was demonstrated
that excess addition of acid in the dissolution step will result in inhibited extraction of rare earth
elements (La, Ce, Pr, Nd) and aluminum. Controlling the acidity of the produced leachate solution is
therefore important, because extracting agent is consumed through preferential extraction of acid
[114].
6.3 Recovery of reusable AB5 alloy

This work set out to test carboxylic acids for the recovery of reusable AB5 alloy, as had been previously
proposed [115]. The results from this investigation indicated that the chemical separation of cathode
and anode (including AB5 alloy) was possible at 50 °C, using 1 M maleic acid for 60 min. However, much
(50%) of the AB5 alloy was also dissolved. Therefore, for practical recycling applications the uses of
carboxylic (malonic, maleic, acetic and citric) acids are limited. Avoiding oxidation of AB5 alloy in aqueous
solutions is difficult, due to the AB5 alloys electrochemical reactions with acid and water.
50
Another intention of using carboxylic acids was to slow down the dissolution of the anode material, in
order to remove corrosion products without dissolving significant amount of the AB5 bulk. The
conclusions drawn from dissolution experiments as well as from XRD and SEM analyses were that it is
possible to selectively remove REE(OH)3 corrosion products using acetic or citric acid.
6.4 Recovery of cobalt

The aim of using ascorbic acid was to investigate if cobalt could be recovered selectively from cobalt-
enriched surface layers on cathode nickel hydroxide particles. In a relatively short time (60 min) the
dissolution of these layers was complete. The pH of the solutions did not influence the dissolution rate.
Selective dissolution of cobalt-enriched surface layers is possible because the cobalt oxidation state(s)
are higher than 2 (+2.7, +3 or a mix thereof), which is different compared to those of the other metallic
or hydroxide components.
At a high S/L ratio (0.5 kg cathode material/L 1 M ascorbic acid solution) a concentrated leachate
solution (249 mmol/L Co) was produced. It was also found that precipitation of cobalt, nickel, zinc and
manganese compounds occurred in concentrated leachate solutions.
An interesting result was that manganese contamination, which had been detected in the spent cathode
materials, was indicated to exist in the outermost layers of the nickel hydroxide particles. Complete
dissolution of manganese was obtained after 40 min, in 1 M ascorbic acid solution.
The spent anode material was not affected by the ascorbic acid treatment, and therefore this treatment
can be used as a pre-step in a method to recover useable AB5 alloy.
51
7. FUTURE WORK
The work presented in the current thesis has suggested three different methods of extracting different
valuable materials (Ni(s), AB5 alloy and Co) from spent NiMH electrodes. These three methods are
interesting to further investigate in order to develop and/or optimize practical recycling methods.
The feasibility of separating particles of metallic nickel from mixed electrode waste was demonstrated
using hydrochloric acid. For recycling purposes, it is necessary to further investigate a practical method
to control the amount of added acid. It is important to either completely dissolve the solid materials or
control the acidity of the leachate solution. Addition of oxidation agents, such as hydrogen peroxide,
may be useful to achieve complete Ni(s) dissolution. The effect of added chemicals to subsequent
separation and steps needs to be considered. Controlling the acidity would be another alternative,
which involves developing a practical method of measuring the rate of acid consumption.
Recovery of AB5 alloy using aqueous method is challenging due to oxidation of AB5 alloy through
hydrogen reduction from water or acid. It is necessary to inhibit the surface oxidation. Nitrogen
containing compounds (i.e. amines) have a high base strength and the free electron pairs on nitrogen
can associate and bind to the metal surface. Potentially, amines can inhibit oxidation more than similar
oxygen-containing compounds (i.e. carboxylic acids).
To separate cobalt using ascorbic acid in practical recycling it is recommended that further
investigations of this system focus on the aqueous system perspectives (i.e. complex solubility). This
recommendation is made due to observations of precipitation of cobalt, nickel, zinc and manganese in
some leachate solutions, while attempting to scale up this method. It is also recommended that further
investigations are designed to further evaluate the type of cobalt complex that is formed, in order to
develop leachate treatment methods, e.g. solvent extraction.
52
8. ACKNOWLEDGEMENTS
There are many that have contributed to this work and the author would like to thank:
The Swedish Energy Agency for funding this work (project P37720-1: Process development for reuse
and/or recycling of Nickel metal hydride batteries).
The battery manufacturer Nilar AB, through Stina Starborg and Erika Widenkvist Zetterström, for
good cooperation, introduction to battery chemistry and production, and supplying batteries and
materials.
Supervisor Britt-Marie Steenari for all her time, patience, support and guidance.
Assistant supervisors Martina Petranikova and Burcak Ebin for useful ideas and support through
practical work.
The Examiner (Prof. Christian Ekberg) and Reviewers.
Stellan Holgersson for help with BET measurements.
Anders Kvist, Stefan Gustafsson and Britt-Marie for SEM introduction.
Britt-Marie and Burcak for help with XRD measurements.
Yu Cao for performing XPS analysis and data interpretations.
Rikard Ylmén for help with AAS, suggestions and computational contributions.
Fredrik Nyhlén for your good lab work and interesting ideas.
My current and former great colleagues at NC/IMR who helped to make this work possible.
My office-mate Toni Gutknecht for putting up with me and bringing laughter to the office space.
My whole family and all my friends who made this work possible.
My dear girlfriend Elin Hidebring for all your love and support.
Thank You All,

Filip Holmberg
Gothenburg, Sweden
May 2017
53
9. BIBLIOGRAPHY
1. Köhler, U., C. Antonius, and P. Bäuerlein, Advances in alkaline batteries. Journal of Power
Sources, 2004. 127(1-2): p. 45-52.
2. Biendicho, J.J., et al., In situ investigation of commercial Ni(OH)2 and LaNi5-based electrodes by
neutron powder diffraction. Journal of Materials Research, 2015. 30(3): p. 407-416.
3. Sato, N. and K. Yagi, Thermal behavior analysis of nickel metal hydride batteries for electric
vehicles. JSAE review, 2000. 21(2): p. 205-211.
4. Taniguchi, A., et al., Development of nickel/metal-hydride batteries for EVs and HEVs. Journal of
Power Sources, 2001. 100(1-2): p. 117-124.
5. Taheri, P., M. Yazdanpour, and M. Bahrami, Analytical assessment of the thermal behavior of
nickel–metal hydride batteries during fast charging. Journal of Power Sources, 2014. 245: p.
712-720.
6. Battery module produced at Nilar AB. www.nilar.com.
7. The European Commisson. Directive 2006/66/EC of the european parliament and of the council.
2006 [cited 2015 02-10]; Available from: ec.europa.eu.
8. European Rare Earths Compentency Network. Strengthening the european rare earths supply-
chain. 2015; Available from: http://ec.europa.eu/.
9. Umicore. Available from: www.umicore.com.
10. Nickel-Hütte Aue GMBH. Available from: http://www.nickelhuette.com/.
11. Lin, S.-L., et al., Characterization of spent nickel–metal hydride batteries and a preliminary
economic evaluation of the recovery processes. Journal of the Air & Waste Management
Association, 2016. 66(3): p. 296-306.
12. Shanghai Metals Market. 2017-02-10]; Available from: www.metal.com.
13. Eckelman, M.J., Facility-level energy and greenhouse gas life-cycle assessment of the global
nickel industry. Resources, Conservation & Recycling, 2010. 54(4): p. 256-266.
14. The European Commisson, Report on critical raw materials for the EU. 2014: ec.europa.eu.
15. U.S. Department of Energy. Critical Materials Strategy. 2010 [cited 2016 11-12]; Available from:
energy.gov.
16. Binnemans, K., et al., Recycling of rare earths: a critical review. Journal of Cleaner Production,
2013. 51(0): p. 1-22.
17. Alonso, E., et al., Evaluating rare earth element availability: A case with revolutionary demand
from clean technologies. Environmental Science and Technology, 2012. 46(6): p. 3406-3414.
18. Sullivan, J.L. and L. Gaines, Status of life cycle inventories for batteries. Energy Conversion and
Management, 2012. 58: p. 134-148.
19. A review of battery life-cycle analysis: State of knowledge and critical needs. 2011. p. 91-133.
20. Majeau-Bettez, G., T.R. Hawkins, and A.H. StrØmman, Life cycle environmental assessment of
lithium-ion and nickel metal hydride batteries for plug-in hybrid and battery electric vehicles.
Environmental Science and Technology, 2011. 45(10): p. 4548-4554.
21. Ye, Z., et al., Metal hydrides for high-power batteries. MRS Bulletin, 2013. 38(6): p. 504-508.
22. Ye, Z., et al., Oxygen and hydrogen gas recombination in NiMH cells. Journal of Power Sources,
2012. 208: p. 232-236.
23. Bäuerlein, P., et al., Progress in high-power nickel–metal hydride batteries. Journal of Power
Sources, 2008. 176(2): p. 547-554.
24. Reddy, T., Linden's Handbook of Batteries (4th Edition). 2010: McGraw-Hill Professional
Publishing.
25. Fetcenko, M.A., et al., Recent advances in NiMH battery technology. Journal of Power Sources,
2007. 165(2): p. 544-551.
26. Encyclopædia Britannica. lanthanoid contraction. 2011 [cited 2017 02-15]; Available from:
https://global.britannica.com/science/lanthanoid-contraction.
27. Kleperis, J., et al., Electrochemical behavior of metal hydrides. Journal of Solid State
Electrochemistry, 2001. 5(4): p. 229-249.
54
28. Percheron-Guégan, A. and J.-M. Welter, Preparation of intermetallics and hydrides. Hydrogen
in Intermetallic Compounds I, ed. L. Schlapbach. 1988: Springer.
29. Schlapbach, L., et al., Surface effects and the formation of metal hydrides. Journal of The Less-
Common Metals, 1980. 73(1): p. 145-160.
30. Khaldi, C., et al., Corrosion effect on the electrochemical properties of LaNi
3.55Mn0.4Al0.3Co0.75 and LaNi 3.55Mn0.4Al0.3Fe0.75 negative electrodes used in Ni-MH
batteries. Materials Science and Engineering B: Solid-State Materials for Advanced Technology,
2010. 175(1): p. 22-28.
31. MCBreen, J. and J.J. Reilly, Synthesis and characterization of metal hydride electrodes. 1995,
Division of chemical sciences: New York.
32. Bode, H., K. Dehmelt, and J. Witte, Zur kenntnis der nickelhydroxidelektrode—I. Über das nickel
(II)-hydroxidhydrat1. Electrochimica Acta, 1966. 11(8): p. 1079-1087.
33. Hall, D.S., et al., Nickel hydroxides and related materials: a review of their structures, synthesis
and properties. PROCEEDINGS OF THE ROYAL SOCIETY A-MATHEMATICAL PHYSICAL AND
ENGINEERING SCIENCES, 2015. 471(2174): p. 20140792-20140792.
34. Bode, H., K. Dehmelt, and J. Witte, Zur kenntnis der nickelhydroxidelektrode—I.Über das nickel
(II)-hydroxidhydrat. Electrochimica Acta, 1966. 11(8): p. 1079-IN1.
35. Besenhard, J.O., Handbook of battery materials. 1999, Weinheim: Wiley-VCH.
36. Chen, J., et al., The effect of Zn(OH)(2) addition on the electrode properties of nickel hydroxide
electrodes. Journal of Materials Research, 1999. 14(5): p. 1916-1921.
37. Leblanc, P., MECHANISM OF ALLOY CORROSION AND CONSEQUENCES ON SEALED NICKEL-
METAL HYDRIDE BATTERY PERFORMANCE. Journal of the Electrochemical Society, 1998. 145(3):
p. 860-863.
38. Li, L., F. Wu, and K. Yang, Degradation analysis of nickel/metal hydride battery and its electrode
materials. Trans. Nonferrous Met. Soc. China, 2004. 14(3).
39. Shinyama, K., et al., Deterioration mechanism of nickel metal-hydride batteries for hybrid electric
vehicles. Journal of Power Sources, 2005. 141(1): p. 193-197.
40. Mavis, B., Homogeneous Precipitation of Nickel Hydroxide Powders. 2009.
41. Liu, L.P., Z.T. Zhou, and C.H. Peng, Sonochemical intercalation synthesis of nano gamma-nickel
oxyhydroxide: Structure and electrochemical properties. Electrochimica Acta, 2008. 54(2): p.
434-441.
42. Hu, B., et al., Controllable Synthesis of Zinc‐Substituted α‐ and β‐Nickel Hydroxide
Nanostructures and Their Collective Intrinsic Properties. Chemistry – A European Journal, 2008.
14(29): p. 8928-8938.
43. Dou, S.X., J.B.D.H. Chen, and H.K. Liu, Nickel hydroxide as an active material for the positive
electrode in rechargeable alkaline batteries. Journal of the Electrochemical Society, 1999.
146(10): p. 3606-3612.
44. Pickett, D.F. and J.T. Maloy, MICROELECTRODE STUDIES OF ELECTROCHEMICALLY
COPRECIPITATED COBALT HYDROXIDE IN NICKEL HYDROXIDE ELECTRODES. Journal of the
Electrochemical Society, 1978. 125(7): p. 1026-1032.
45. Ratke, L., et al., Growth and Coarsening: Ostwald Ripening in Material Processing. Vol. 1. 2002,
Berlin, Heidelberg: Springer Berlin Heidelberg.
46. Peng, M.-x., et al., Template growth mechanism of spherical Ni(OH)2. Journal of Central South
University of Technology, 2007. 14(3): p. 310-314.
47. Van Bomme, A. and J.R. Dahn, Analysis of the growth mechanism of coprecipitated spherical and
dense nickel, manganese, and cobalt-containing hydroxides in the presence of aqueous
ammonia. Chemistry of Materials, 2009. 21(8): p. 1500-1503.
48. Chang, Z., et al., Influence of preparation conditions of spherical nickel hydroxide on its
electrochemical properties. Journal of Power Sources, 1998. 74(2): p. 252-254.
49. Tsai, I.C., Y.Y. Wang, and C.C. Wan, Effect of synthesis method on the properties of Ni(OH)(2) for
Ni/MH batteries. JOURNAL OF NEW MATERIALS FOR ELECTROCHEMICAL SYSTEMS, 2001. 4(2):
p. 99-106.
55
50. Shangguan, E.B., et al., Comparative structural and electrochemical study of high density
spherical and non-spherical Ni(OH)(2) as cathode materials for Ni-metal hydride batteries.
Journal of Power Sources, 2011. 196(18): p. 7797-7805.
51. Shangguan, E., et al., Effects of different Ni(OH)(2) precursors on the structure and
electrochemical properties of NiOOH. International Journal of Hydrogen Energy, 2011. 36(16):
p. 10057-10064.
52. Li, L., et al., Recovery of Ni, Co and rare earths from spent Ni–metal hydride batteries and
preparation of spherical Ni(OH)2. Hydrometallurgy, 2009. 100(1–2): p. 41-46.
53. Zhang, W., et al., Effect of nickel hydroxide composition on the electrochemical performance of
spherical Ni(OH)2 positive materials for Ni–MH batteries. International Journal of Hydrogen
Energy, 2009. 34(1): p. 473-480.
54. Oshitani, M., et al., Development of a pasted nickel electrode with high active material
utilization. Journal of the Electrochemical Society, 1989. 136(6): p. 1590-1593.
55. Ying, T., Surface modification of nickel hydroxide particles by micro-sized cobalt oxide hydroxide
and properties as electrode materials. Surface & Coatings Technology, 2005. 200(7): p. 2376-
2379.
56. Chang, Z., H. Tang, and J.G. Chen, Surface modification of spherical nickel hydroxide for nickel
electrodes. Electrochemistry Communications, 1999. 1(11): p. 513-516.
57. Wang, X.Y., et al., Electrochemical characteristics of nickel hydroxide modified by electroless
cobalt coating. International Journal of Hydrogen Energy, 1998. 23(10): p. 873-878.
58. Wang, X., et al., Surface modification and electrochemical studies of spherical nickel hydroxide.
59. Benson, P., G.W.D. Briggs, and W.F.K. Wynne-Jones, The cobalt hydroxide electrode-I. Structure
and phase transitions of the hydroxides. Electrochimica Acta, 1964. 9(3): p. 275-280.
60. Ye, Z., et al., Ni-MH battery electrodes made by a dry powder process. Journal of the
Electrochemical Society, 1995. 142(12): p. 4045-4050.
61. Naboychenko, S.S., I.B. Murashova, and O.D. Neikov, Production of Cobalt and Cobalt-Alloy
Powders. 2009, Elsevier Science & Technology. p. 399-408.
62. Capus, J.M., Chapter 5 - Technical overview — Metal powder production, in Metal Powders
(Fourth Edition), J.M. Capus, Editor. 2005, Elsevier Science: Oxford. p. 153-165.
63. Lancashire, R.J. Nickel Chemistry. 2010 [cited 2015 02-15]; Available from:
wwwchem.uwimona.edu.jm/courses/nickel.html.
64. Mond, L., C. Langer, and F. Quincke, Action of carbon monoxide on nickel. Journal of the
Chemical Society, Transactions, 1890. 57: p. 749.
65. Possible selective solar photothermal absorber: Ni dendrites formed on Al surfaces by the CVD
of Ni (Co) 4. (USA). Vacuum, 1978. 28(8): p. 377-377.
66. Beattie, J.K., A.F. Masters, and J.T. Meyer, Nickel carbonyl cluster complexes. Polyhedron, 1995.
14(7): p. 829-868.
67. Monnier, J., et al., Identification of a new pseudo-binary hydroxide during calendar corrosion of
(La, Mg).sub.2Ni.sub.7-type hydrogen storage alloys for Nickel-Metal Hydride batteries. Journal
of Power Sources, 2014. 266: p. 162.
68. Ruiz, F.C., et al., Effect of electrolyte concentration on the electrochemical properties of an AB5-
type alloy for Ni/MH batteries. International Journal of Hydrogen Energy, 2013. 38(1): p. 240-
245.
69. Zhou, X., et al., Degradation mechanisms of high-energy bipolar nickel metal hydride battery
with AB5 and A2B7 alloys. Journal of Alloys and Compounds, 2013. 580(1): p. S373-S377.
70. Hu, W.K., et al., Effect of long-term overcharge and operated temperature on performance of
rechargeable NiMH cells. Journal of Power Sources, 2006. 159(2): p. 1478-1483.
71. Ikoma, M., et al., Effect of alkali-treatment of hydrogen storage alloy on the degradation of
Ni/MH batteries. Journal of Alloys and Compounds, 1999. 284(1): p. 92-98.
56
72. Maurel, F., B. Knosp, and M. Backhaus-Ricoult, Characterization of corrosion products of AB5-
type hydrogen storage alloys for nickel-metal hydride batteries. Journal of the Electrochemical
Society, 2000. 147(1): p. 78-86.
73. Iwakura, C., et al., Self-discharge mechanism of nickel-hydrogen batteries using metal hydride
anodes. Journal of the Electrochemical Society, 1989. 136(5): p. 1351-1355.
74. Wang, C.S., et al., Corrosion behaviour of AB5-type hydride electrodes in alkaline electrolyte
solution. Journal of Applied Electrochemistry, 2003. 33(3): p. 325-331.
75. Li Li, W.F., Yang Kai, Wang Jing, Chen., Electrochemical performance of nickel／metal hydride
batteries under unconventional conditions and degradation analysis. Transactions of
Nonferrous Metals Society of China, 2004. 14(1): p. 208-214.
76. Kong, L., et al., Effects of Al- and Mn-contents in the negative MH alloy on the self-discharge and
long-term storage properties of Ni/MH battery. Journal of Power Sources, 2012. 213: p. 128-
139.
77. Shinyama, K., et al., Investigation into the deterioration in storage characteristics of nickel-metal
hydride batteries during cycling. Journal of Power Sources, 2005. 143(1): p. 265-269.
78. Young, K., et al., Microstructures of the oxides on the activated AB2 and AB5 metal hydride alloys
surface. Journal of Alloys and Compounds, 2014. 606: p. 97-104.
79. Sherif, S.A., et al., Handbook of Hydrogen Energy. 2014, Boca Raton: CRC Press.
80. Guerlou-Demourgues, L., C. Denage, and C. Delmas, New manganese-substituted nickel
hydroxides. Journal of Power Sources, 1994. 52(2): p. 269-274.
81. Mantuano, D.P., et al., Analysis of a hydrometallurgical route to recover base metals from spent
rechargeable batteries by liquid-liquid extraction with Cyanex 272. Journal of Power Sources,
2006. 159(2): p. 1510-1518.
82. Pietrelli, L., et al., Characterization and leaching of NiCd and NiMH spent batteries for the
recovery of metals. Waste Management, 2005. 25(2): p. 221-226.
83. Junmin, N., et al., Dismantling, Recovery, and Reuse of Spent Nickel-Metal Hydride Batteries.
Journal of the Electrochemical Society, 2006. 153(1): p. A101-A105.
84. Zhang, P., et al., Recovery of metal values from spent nickel–metal hydride rechargeable
batteries. Journal of Power Sources, 1999. 77(2): p. 116-122.
85. Zhang, P., et al., Hydrometallurgical process for recovery of metal values from spent nickel-metal
hydride secondary batteries. Hydrometallurgy, 1998. 50(1): p. 61-75.
86. Rodrigues, L.E.O.C. and M.B. Mansur, Hydrometallurgical separation of rare earth elements,
cobalt and nickel from spent nickel–metal–hydride batteries. Journal of Power Sources, 2010.
195(11): p. 3735-3741.
87. Wu, F., et al., Recovery of valuable metals from anode material of hydrogen-nickel battery.
Transactions of Nonferrous Metals Society of China, 2009. 19(2): p. 468-473.
88. Pietrelli, L., et al., Rare earths recovery from NiMH spent batteries. Hydrometallurgy, 2002.
66(1–3): p. 135-139.
89. Fernandes, A., J.C. Afonso, and A.J.B. Dutra, Separation of nickel(II), cobalt(II) and lanthanides
from spent Ni-MH batteries by hydrochloric acid leaching, solvent extraction and precipitation.
Hydrometallurgy, 2013. 133(0): p. 37-43.
90. Tzanetakis, N. and K. Scott, Recycling of nickel–metal hydride batteries. I: Dissolution and solvent
extraction of metals. Journal of Chemical Technology & Biotechnology, 2004. 79(9): p. 919-926.
91. Innocenzi, V. and F. Vegliò, Recovery of rare earths and base metals from spent nickel-metal
hydride batteries by sequential sulphuric acid leaching and selective precipitations. Journal of
Power Sources, 2012. 211: p. 184-191.
92. Larsson, K., C. Ekberg, and A. Ødegaard-Jensen, Dissolution and characterization of HEV NiMH
batteries. Waste Management, 2013. 33(3): p. 689-698.
93. Larsson, K., et al., Using Cyanex 923 for selective extraction in a high concentration chloride
medium on nickel metal hydride battery waste: Part II: Mixer-settler experiments.
Hydrometallurgy, 2013. 133: p. 168-175.
57
94. Provazi, K., et al., Metal separation from mixed types of batteries using selective precipitation
and liquid–liquid extraction techniques. Waste Management, 2011. 31(1): p. 59-64.
95. Tzanetakis, N. and K. Scott, Recycling of nickel–metal hydride batteries. II: Electrochemical
deposition of cobalt and nickel. Journal of Chemical Technology & Biotechnology, 2004. 79(9):
p. 927-934.
96. Innocenzi, V. and F. Veglio, Separation of manganese, zinc and nickel from leaching solution of
nickel-metal hydride spent batteries by solvent extraction. Hydrometallurgy, 2012. 129-130: p.
50-58.
97. Nayl, A.A., Extraction and separation of Co(II) and Ni(II) from acidic sulfate solutions using
Aliquat 336. Journal of hazardous materials, 2010. 173(1): p. 223-230.
98. Juang, R.-S. and H.-C. Kao, Extraction separation of Co(II)/Ni(II) from concentrated HCl solutions
in rotating disc and hollow-fiber membrane contactors. Separation and Purification Technology,
2005. 42(1): p. 65-73.
99. Rice, N., H. Irving, and M. Leonard, Nomenclature for liquid-liquid distribution (solvent
extraction)(IUPAC Recommendations 1993). Pure and applied chemistry, 1993. 65(11): p. 2373-
2396.
100. Bard, A.J., et al., Electrochemical Dictionary. 1. Aufl. ed. 2008, Berlin, Heidelberg: Springer Berlin
Heidelberg.
101. Lupi, C. and D. Pilone, Ni–MH spent batteries: a raw material to produce Ni–Co alloys. Waste
Management, 2002. 22(8): p. 871-874.
102. Nan, J., et al., Recovery of metal values from a mixture of spent lithium-ion batteries and nickel-
metal hydride batteries. Hydrometallurgy, 2006. 84(1–2): p. 75-80.
103. Dixini, P.V.M., et al., Recycling of the anode from spent Ni-MH batteries for synthesis of
the lanthanide oxysulfide/oxysulfate compounds used in an oxygen storage and release system.
104. Bertuol, D.A., A.M. Bernardes, and J.A.S. Tenório, Spent NiMH batteries—The role of selective
precipitation in the recovery of valuable metals. Journal of Power Sources, 2009. 193(2): p. 914-
923.
105. Jones, D.A., Principles and prevention of corrosion. Vol. 2. 1996, Englewood Cliffs, N.J: Prentice
Hall.
106. Atkins, P.W. and L. Jones, Chemical principles: the quest for insight. Vol. 4. 2008, New York: W.H.
Freeman.
107. Vanýsek, P., CRC Handbook of Chemistry and Physics. 87 ed. 2006, Boca Raton, Florida: CRC
Press.
108. Christmann, K., et al., Adsorption of hydrogen on nickel single crystal surfaces. The Journal of
Chemical Physics, 1974. 60(11): p. 4528-4540.
109. Konvalinka, J., P. Vanoeffelt, and J. Scholten, Temperature programmed desorption of hydrogen
from nickel catalysts. Applied Catalysis, 1981. 1(3-4): p. 141-158.
110. McCafferty, E., Introduction to corrosion science. 2010, New York: Springer.
111. Speight, J., et al., Lange's Handbook of Chemistry. Vol. 16th,Anniversary,Special,Revis;16th;.
2004, Blacklick;New York;: McGraw-Hill Professional Publishing.
112. Rydberg, J., ed. Solvent extraction principles and practice. 2004, Marcel Dekker Inc.
113. Larsson, K., Hydrometallurgical treatment of NiMH batteries. 2012, Chalmers University of
Technology: Göteborg.
114. Alguacil, F.J. and F.A. López, The extraction of mineral acids by the phosphine oxide Cyanex 923.
Hydrometallurgy, 1996. 42(2): p. 245-255.
115. Smith, W.N. and S. Swoffer, Process for the Recovery of AB5 Alloy from Used Nickel/Metal
Hydride Batteries. 2014, Google Patents.
116. Martell, A.E. and R.M. Smith, Critical stability constants: Vol. 3, Other organic ligands. 1977, Ney
York: Plenum.
58
117. Amin, M.A., H. Shokry, and E.M. Mabrouk, Nickel Corrosion Inhibition in Sulfuric Acid-
Electrochemical Studies, Morphologies, and Theoretical Approach. Corrosion, 2012. 68(8): p.
699.
118. Ferreira, E.S., et al., Evaluation of the inhibitor effect of l-ascorbic acid on the corrosion of mild
steel. Materials Chemistry and Physics, 2004. 83(1): p. 129-134.
119. Amin, M.A., et al., The inhibition of low carbon steel corrosion in hydrochloric acid solutions by
succinic acid. Electrochimica Acta, 2007. 52(11): p. 3588-3600.
120. Tizpar, A. and Z. Ghasemi, The corrosion inhibition and gas evolution studies of some surfactants
and citric acid on lead alloy in 12.5 M H2SO4 solution. Applied Surface Science, 2006. 252(24):
p. 8630-8634.
121. Du, J., J.J. Cullen, and G.R. Buettner, Ascorbic acid: Chemistry, biology and the treatment of
cancer. Biochimica et Biophysica Acta - Reviews on Cancer, 2012. 1826(2): p. 443-457.
122. Borsook, H. and G. Keighley, Oxidation-reduction Potential of Ascorbic Acid (Vitamin C).
Proceedings of the National Academy of Sciences of the United States of America, 1933. 19(9):
p. 875-878.
123. Buettner, G.R. and B.A. Jurkiewicz, Catalytic Metals, Ascorbate and Free Radicals: Combinations
to Avoid. Radiation Research, 1996. 145(5): p. 532-541.
124. Li, L., et al., Ascorbic-acid-assisted recovery of cobalt and lithium from spent Li-ion batteries.
Journal of Power Sources, 2012. 218: p. 21-27.
125. Casas-Cabanas, M., Rodriguez-Carvajal, J., Canales-Vazquez, J., Palacin, M. J., New insights on
the microstructural characterization of nickel hydroxides and correlation with electrochemical
properties. Mater. Chem. 16, 2925 (2006).
126. Construction et mise a point d`une chambre de precision pour siagrammes de rayons X a haute
temperature. Diament, R. Met. Corros.-Ind. 31, 167 (1956).
127. Chartouni D., K.N., Otto A., Guther V., Nutzenader C., Zuttel A., Influence of the alloy morphology
on the kinetics of AB5-type metal hydride electrodes. Alloys Compd. 293, 292 (1999).
128. Mullica, D.F., Oliver, J.D., Milligan, W.O., Cerium Trihydroxide. Acta Crystallogr. Sec. B: Struct.
Crystallogr. Cryst. Chem. 35, 2668 (1979).
129. Bomben, K.D., et al., Handbook of x-ray photoelectron spectroscopy. A reference bookof
standard spectra for identification and interpretation of xpsdata. 1995, Eden Prairie, MN:
Physical Electronics.
130. Grosvenor, A.P., et al., New interpretations of XPS spectra of nickel metal and oxides. Surface
Science, 2006. 600(9): p. 1771-1779.
131. Joint Committee of Powder Diffraction Standards, JCPDS-ICCD, Editor. 2010: Philadelphia, USA.
132. Brunauer, S., P.H. Emmett, and E. Teller, Adsorption of gases in multimolecular layers. Journal
of the American Chemical Society, 1938. 60(2): p. 309-319.
59
APPENDIX I: ANALYSIS TECHNIQUES
I. A Qualitative analysis of crystalline compounds by X-ray diffraction (XRD)
Solid samples were analyzed using X-ray diffraction (XRD) instrument Bruker AXS. The samples were
exposed to characteristic Cu Kα irradiation and the diffracted X-rays were detected with a scintillation
detector. Identification of crystalline compounds was made by comparison with standards in the Joint
Committee of Powder Diffraction Standard database [131].
I. B Scanning electron microscope (SEM) imaging of particles

Imaging of solid sample surfaces was carried out using an Ultra 55 field emission gun (FEG) scanning
electron microscope (SEM) and Quanta 200 FEG environmental SEM (ESEM). Both these instruments
were equipped with Oxford Inca energy dispersive X-ray (EDX) detectors. Acceleration voltages were
between 5 and 15 kV.
I. C Elemental composition using inductively coupled plasma optical emission spectroscopy (ICP-OES)
An inductively coupled plasma optical emission spectrometer (iCAP6500) was used for determining
concentrations of elements in solutions. Samples were diluted in 0.7 M suprapure nitric acid (65 wt%,
Sigma Aldrich) and compared to external calibration, prepared from standard stock solutions (1000
ppm, Ultra Scientific). The limit of quantification was determined as 10σ of the standard deviation of
the blank (matrix) solution.
I. D Solid sample surface area measurement using Brunauer-Emmett-Teller (BET) method

The specific surface area (m2/g) of samples was measured using the Brunauer-Emmett-Teller (BET)
method [132]. Nitrogen gas was used for adsorption determination, kept isothermal at 77 K using liquid
nitrogen, using a Micrometrics ASAP 2020. The temperature was ramped with 10 °C/min to 100 °C, and
kept at this temperature for 60 min under vacuum (300 µmHg).
I. E X-ray photoelectron spectroscopy (XPS) analysis of surface species

The X-ray photoelectron spectroscopy (XPS) analyses were carried out using a PHI 5000 VersaProbe III
instrument. The AlKα (1486.6 eV) X-ray source using monochromator was used to generate
photoelectrons. Acquisition conditions for the survey spectra (0-1100 eV) were 140 eV pass energy, 45
take off angle with 0.5 eV/step. Selected region spectra were recorded covering the Ni 2p3/2, Co 2p3/2
and C 1s photoelectron peaks. The acquisition conditions were then 26 or 55 eV pass energy, 45 take
off angle and 0.1 eV/step. The samples were mounted on a removable tape and flooded with low energy
electrons and argon ions, in order to compensate for charge. C 1s (284.8 eV) was used for reference.
The depth profiling was carried out by successive ion etchings and XPS analyses using Ni 2p3/2 and
Co2p1/2 peaks, using an argon beam of 2kV over an area of 2x2 mm 2. The etching rate calibrated on
Ta2O5 with these conditions is 4.76 nm/min.
60

Recycling of Nickel Metal Hydride (Nimh) Batteries: Characterization and Recovery of Nickel, Ab Alloy and Cobalt

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Recycling of Nickel Metal Hydride (Nimh) Batteries: Characterization and Recovery of Nickel, Ab Alloy and Cobalt

Uploaded by

Copyright:

Available Formats

THESIS FOR THE DEGREE OF LICENTIATE OF ENGINEERING

Recycling of nickel metal hydride (NiMH) batteries

Industrial Materials Recycling

© FILIP HOLMBERG, 2017

Licentiatuppsatser vid Institutionen för kemi och kemiteknik

Industrial Materials Recycling

Industrial Materials Recycling

This thesis is based on the work contained in the following paper:

2.2 Fundamentals of the NiMH battery

MH + OH- ⇄ M + H2O + e- Eq. 1 (anode) E° = -0.83 V

NiOOH + H2O + e- ⇄ Ni(OH)2 + OH- Eq. 2 (cathode) E° = 0.52 V

MH + NiOOH ⇄ M + Ni(OH)2 Eq. 3 (total cell reaction) E° = 1.35 V

2.2.1 Hydrogen storage AB5 alloy used in the anode

2.2.2 Nickel hydroxide Ni(OH)2 used in the cathode

To manufacture a spherical form of nickel hydroxide to be used in NiMH batteries it is necessary to

Temp. Chem. time Drying Size Tapping density

2.2.3 Metallic powders nickel and cobalt

2.3 Degradation mechanisms of battery cells and materials

2.3.4 Cross contamination between cathode and anode

Anode, cathode 2 M H2SO4 20 2h 1/10 76.6% 97.6% > 90% [82]

Mixed 3 M H2SO4 75 5h 1/10 > 95% >95% > 95% [83]

Mixed 2 M H2SO4 95 4h 5/100 97% 100% 96% [84]

Mixed 3 M HCl 95 3h 1/10 97.4% 100% 99.1% [85]

Mixed 1.5 M H2SO4 30 1h 1/20 97% 95-100% 87% [86]

Anode 3 M H2SO4 95 4h 1/10 99.5% 99.5% 94.5% [87]

Mixed 2 M H2SO4 20 2h 1/10 > 76.6% >97.6% > 91.8% [88]

Mixed 12 M HCl 40 2h 15/100 n/a n/a n/a [89]

Mixed 4 M HCl 95 3h 1/10 98% 100% 99% [90]

Mixed 4 M H2SO4 80 3h n/a 100% 98% 36% [91]

Mixed 8 M HCl 30 n/a 3.7/10 n/a 100% 100% [93]

n/a – not available or not presented in detail.

2.4.2 NiMH metal separation by solvent extraction

𝑁𝑖(𝑠) ⇄ 𝑁𝑖 2+ (𝑎𝑞) + 2𝑒 − Eq. 5

2𝐻 + (𝑎𝑞) + 2𝑒 − ⇄ 𝐻2 (𝑔) Eq. 6

= 0 𝑉 + 0.26 𝑉 − 0.236 𝑉 = 0.024 𝑉

3.2 Recovery of nickel

𝑁𝑖(𝑂𝐻)2 + 2𝐻𝐶𝑙 (𝑎𝑞) → 𝑁𝑖𝐶𝑙2 (𝑎𝑞) + 2𝐻2 𝑂(𝑙) Eq. 9

𝑁𝑖(𝑠) + 2𝐻𝐶𝑙 (𝑎𝑞) → 𝑁𝑖𝐶𝑙2 (𝑎𝑞) + 𝐻2 (𝑔) Eq. 10

3.2.2 Distribution into the organic phase in solvent extraction

3.2.3 Extraction of chloride complexes with Cyanex 923

𝑀 𝑥+ (𝑎𝑞) + 𝑥𝐶𝑙 − (𝑎𝑞) + 𝑦𝐵(𝑜𝑟𝑔) ⇄ 𝑀𝐶𝑙𝑥 𝐵𝑦 (𝑜𝑟𝑔) Eq. 11

𝐻 + (𝑎𝑞) + 𝐶𝑙 − (𝑎𝑞) + 𝐵(𝑜𝑟𝑔) ⇄ 𝐻𝐶𝑙𝐵(𝑜𝑟𝑔) Eq. 12

3.3 Recovery of reusable anode using carboxylic acids

R-(COOH)x + xH2O = R-(COO -)x + xH3O+ Eq. 13

(1) Characterization of unspent and spent cathode and anode.

(2) Recovery of Ni(s) from the conductive nickel network.

4.2 Battery disassembly

Leachate treatment Metal extraction by variation of O:A

4.4.1 Dissolution rates of metals at pH 1

4.4.2 Nickel dissolution rates at 1-8 M HCl

4.4.3 Leachate treatment with solvent extraction

4.5 Recovery of anode material (including AB5 alloy)

Characterization of undissolved samples containing anode components is important for the

4.5.1 Dissolution experiments using carboxylic acids

4.6.1 Dissolution experiments using ascorbic acid

Table 13. Composition of spent cathode and anode electrode samples.

[a] Mixed electrodes [b] Mixed electrodes

Volume added (mL)

Undissolved material after 5 h of dissolution of mixed electrode materials in hydrochloric acid at pH 1

Metal dissolution (%)

Pure Ni(s) powder

8 vol% Cyanex 923

70 vol% Cyanex 923